Vanilla DramaBox TTS — Generate + Reward-Rank Benchmark

We took two-scene "CUT TO:" DramaBox stage-direction prompts (a voice/style description, a first scene of dialogue, a hard CUT TO:, then a second scene in a contrasting emotion) and ran them through the MOSS-TTS 8B voice-acting model with no reference audio. For every prompt we generate 4 different takes (4 random seeds), automatically score each take, and keep the best. This page reports how fast that is, how much quality you lose if you generate fewer than 4 takes, and what it would cost to scale to 10k / 1M prompts.

These are the RAW, vanilla DramaBox outputs. Every two-scene CUT TO: DramaBox prompt was fed directly to the 8B MOSS voice-acting TTS (the quoted dialogue → text to say, the stage-direction → style instruction), sampled with 4 seeds, then scored and reward-ranked. No roundtrip, no speech-restoration / Sidon, no enhancement, no re-synthesis — the audio you hear is the model’s raw output. The “8B MOSS model” is simply the TTS engine that produced this audio; that is correct and intended.
How a prompt is fed (the way the existing DramaBox pipeline does it): the spoken dialogue (the quoted lines) is given as the text to say; the surrounding two-scene stage direction (voice description, emotions, CUT TO:) is given as the style instruction with the literal quotes removed — leaving the quotes inside the instruction makes the 8B ramble and never stop (~0.02 clip/s, most takes dropped). Very long dialogues are capped to the first ~85 words so a take fits the model's ~57 s / 720-token window (median spoken length here ≈ 52s of audio).

The reward (how “best take” is decided)

For each group of 4 takes we compute three scores per take and min–max normalize each score WITHIN the group of 4 (so the worst take of the four = 0, the best = 1 on that axis), then add them:

reward = norm(invWER) + norm(genuineness) + norm(blend)

The 3 terms are added, not multiplied. Reward ranges 0–3. Ranking picks the highest-reward take of the 4.

Headline numbers

200
prompts benchmarked (groups of 4)
796
takes generated & scored
24.2s
per group of 4 · 1×A100
0.17
clips / sec / A100
0.45
mean invWER (all takes)
1.38
mean genuineness /6
5.05
mean blend /10

Best-of-4 winner quality by pathway (5 prompt archetypes × EN/DE mix)

PathwayninvWERgenuineness /6blend /10reward /3
ACCC (Acting Challenge)400.651.847.202.47
CC2-C (Archetype)400.531.796.902.44
CCA (VoiceNet)380.761.757.122.58
Extreme Physical400.671.635.932.42
SIT (Situations)380.721.726.862.59

1 · Timing — one group of 4 on a single A100

Measured over 196 real groups (batched: the 4 seeds run in one generate() call, sdpa attention, bf16), average spoken length 71 words ≈ 39.9s of audio per take.

24.2s
mean / group of 4
25.9s
median / group of 4
0.17
clips / s / A100
149
groups / h / A100

Wall-clock vs number of seeds (batch size), fixed 720-token window, single A100

k (seeds/group)group wall-clockclips/sper-clip smarginal s / extra seedpeak VRAM
k=117.6s0.05717.58s
k=222.5s0.08911.25s4.92s
k=325.1s0.1208.36s2.57s
k=425.6s0.1566.41s0.58s
k=830.3s0.2643.79s1.16s26.7 GB
k=1640.5s0.3952.53s1.28s29.2 GB
k=3259.6s0.5371.86s1.19s34.1 GB
Takeaway — decode is latency-bound; large k is cheap per extra seed until compute saturates. At small k, adding seeds is almost free: going k=1→4 adds only ~8s of wall-clock total for 3 extra seeds, and the 4th seed costs just ~0.6s. Autoregressive decoding is latency/bandwidth-bound there, so a group of 4 barely costs more than a group of 2 — fewer seeds barely saves time; it mainly costs quality (next section). Compute begins to saturate around k≈8: from k=8 onward each extra seed settles at a steady ~1.2s and wall-clock grows roughly linearly (30.3s→40.5s→59.6s for k=8/16/32), while throughput keeps climbing to 0.54 clips/s at k=32 — 9.4× the k=1 rate. k=32 fits in only ~34 GB, far under an 80 GB A100 (no OOM), so for best-of-k planning seeds up to ~8 are near-free and 8→32 stays cheap and linear.

2 · Best-of-k trade-off (k = 4 vs 3 vs 2)

For each of 196 full groups we pick the top-reward take among the first k seeds (fixed group-of-4 normalization). Below: the average winner quality at each k, and how much is LOST going from 4 seeds down to 3 or 2.

keep best ofmean reward /3mean invWERmean genuineness /6mean blend /10
k=42.4990.6661.7476.798
k=32.3290.6201.6966.585
k=22.1160.5711.6276.205

Lost vs best-of-4

vs best-of-4Δ rewardΔ invWERΔ genuinenessΔ blend
only first 3 seeds−0.170−0.046−0.051−0.212
only first 2 seeds−0.383−0.095−0.120−0.593
Interpretation: dropping from 4 takes to 2 loses ~0.38 reward, 9.5 pts of invWER and 0.12 genuineness on average — the classic best-of-N diminishing-returns curve. Since generating 4 vs 2 seeds costs almost the same wall-clock (section 1), k=4 is cheap insurance.

3 · Scaling estimates

Assumptions from the measurements above (1×A100): group-of-4 ≈ 25.6s, group-of-3 ≈ 25.1s, group-of-2 ≈ 22.5s; prompt-gen (gemma-3n-E4B bf16, vLLM) ≈ 20,000/h. Multi-GPU scales ~linearly (independent shards).

DramaBox TTS for 10,000 groups — wall-clock on 1×A100

k1×A100 wall-clock8×A100 (linear)winner mean invWERwinner mean genu
k=43.0 days8.9 h0.6661.747
k=32.9 days8.7 h0.6201.696
k=22.6 days7.8 h0.5711.627
The WER/genuineness columns are the measured best-of-k trade-off from section 2 — i.e. going to k=2 for 10k groups saves 8.8 h but costs 9.5 pts invWER and 0.12 genuineness.

Combined pipeline (prompt-gen THEN TTS) — 10,000 texts

setupk=4k=3k=2
ONE GPU (gen prompts, then TTS)3.0 days2.9 days2.6 days
TWO GPUs (prompt-gen ∥ TTS, pipelined)3.0 days2.9 days2.6 days
Prompt-gen for 10k texts is only ~30.0 min (gemma-3n is ~100× faster than TTS per item), so on one GPU it is a small tax on top of TTS. Running prompt-gen and TTS on separate GPUs hides it entirely — TTS dominates and total wall-clock ≈ the TTS time alone.

1,000,000 texts — end-to-end (prompt-gen + DramaBox TTS)

kprompt-gen GPU-hTTS GPU-htotal GPU-hwall-clock on 8×A100
k=4507,1257,17537.4 days
k=3506,9647,01436.5 days
k=2506,2506,30032.8 days
Prompt-gen for 1M texts ≈ 50 GPU-h (~2 GPU-days) — a rounding error next to TTS. At k=4, 1M DramaBox groups ≈ 7,125 GPU-h ≈ 37.4 days on 8×A100; k=2 cuts that to ≈ 32.8 days at the section-2 quality cost. Scale the GPU count linearly for other cluster sizes.

4 · Showcase — prompts & their 4 ranked takes

10 English prompts per pathway plus a few German. Each shows the full two-scene DramaBox prompt and its 4 takes ranked by reward (🏆 = kept). Audio lazy-loads on play.

CCA (VoiceNet) — 12 prompts

EnglishCCA (VoiceNet)group g0080 · 4 takes
A 65 male with a heavily resonant, bass-heavy voice, delivering this high-quality studio voice recording with no background noise. The same voice is deeply ashamed, almost shamefully grateful. (He shifts his weight slightly.) "I... I don't know what to say." (A faint, almost imperceptible sigh.) "This is… unexpected, isn’t it?" (He clears his throat.) "Truly, I am overwhelmed." (A low rumble vibrates through the audio.) "Your kindness… it’s more than I deserve." (A subtle pause, a slight tremor in the voice.) "I keep thinking… did I really do enough?" (He adjusts something subtly.) "No, of course not. Never enough." (A deeper, more resonant tone enters.) "But this… this is a tremendous weight lifted." (A hint of a smile in the vocal timbre.) "Thank you. From the absolute bottom of… everything." (A brief, choked sound.) "You have no idea the difference this makes." (A slight shift in the cadence.) "It’s… it’s a profound relief." (A long, low hum.) "I… I don’t know how to repay such grace." (A final, weighty resonance.) "Just… knowing." (A soft, almost inaudible exhale.) "Knowing." CUT TO: The same voice now filled with cold, unwavering contempt. (His tone sharpens, a thin edge enters.) "So, you think that buys you forgiveness?" (A barely contained surge of power.) "You believe a few empty words absolve you?" (A deliberate pause, a chilling stillness.) "Don't mistake my gratitude for weakness." (The resonance deepens, becoming more menacing.) "I see the calculation in your eyes." (A subtle hardening of the vocal cords.) "The thinly veiled attempt at self-satisfaction." (A sharp intake of breath.) "You offered a gesture. I accept the consequence." (The bass frequency intensifies.) "And believe me, consequences are… thorough." (A slow, deliberate cadence.) "Do not mistake this for gratitude." (A slight, almost imperceptible shift in tone.) "Consider it… acknowledgment." The performance across both moments feels like a chasm opening between worlds.
🏆
reward 2.15invWER 0.74genu 2.46blend 10.00dur 36sseed 3
ASR heard: “I I I don't know what to say. This is unexpected, isn't it? Truly I am overwhelmed. Your kindness, it's more than I deserve. I keep thinking, did I re”
#2
reward 1.87invWER 0.99genu 2.20blend 5.30dur 55sseed 1
ASR heard: “I don't know what to say. This is unexpected, isn't it? Truly, I am overwhelmed. Your kindness, it's more than I deserve. I keep thinking, did I reall”
#3
reward 1.19invWER 0.69genu 2.06blend 8.03dur 55sseed 2
ASR heard: “I I don't know what to say. This is unexpected, isn't it? Truly, I am overwhelmed. Your kindness, it's more than I deserve. I keep thinking, did I rea”
#4
reward 0.48invWER 0.84genu 1.58blend 4.40dur 55sseed 0
ASR heard: “I I don't know what to say. This is unexpected, isn't it? Truly, I am overwhelmed. Your kindness, it's more than I deserve. I keep thinking, did I rea”
EnglishCCA (VoiceNet)group g0081 · 4 takes
A 42 female with a warm, slightly dark alto, delivering this high-quality studio voice recording with no background noise. The same voice is brimming with a sharp, playful energy. (Slight shift in posture) "Honestly, you're still going on about that? It was *one* little thing, and you've been simmering about it for weeks." "Honestly, it's just... remarkable how you can hold a grudge about something so utterly insignificant, don't you think?" "It's almost impressive, really, the sheer amount of mental energy you dedicate to this." "I mean, I've had far more pressing matters on my mind, things that actually require attention, you know?" "Like trying to find a decent cup of coffee this morning – that was a genuine crisis." "But no, this is far more important, apparently. This deserves all your focus." "It's quite a testament to your dedication, this persistent little cloud of… well, whatever this is." "You’ve really mastered the art of the sustained, low-level annoyance, haven’t you?" (Slight pause, a breath) "It’s a talent, really." "You could probably write a whole book on petty grievances, you know." "A fascinating study in human stubbornness, perhaps." "I almost feel bad for the poor book." "Almost." CUT TO: The same voice now fragile and filled with a quiet sorrow. (Voice breaking) "And it's just… gone. Everything. Just… vanished." "I keep reaching for it, you know? Like it’s still there, just out of reach." "It was so vibrant, so full of… light. Now there’s just this hollow space." "I try to remember the way it felt, the way it sounded, but it’s fading." "Like a photograph left in the sun, slowly bleaching." "I keep expecting to see it, to hear it again, just around the corner." "But there’s nothing. Just emptiness." "It’s a strange kind of ache, this absence." "A constant, dull throb beneath everything else." "I miss it so much." (Quiet sigh) "I just… miss it." The performance across both moments feels [a stark contrast between sharp-edged amusement and profound vulnerability].
🏆
reward 3.00invWER 1.00genu 2.18blend 5.74dur 27sseed 2
ASR heard: “Honestly, you're still going on about that? It was one little thing, and you've been simmering about it for weeks. Honestly, it's just remarkable how ”
#2
reward 1.87invWER 0.99genu 1.00blend 4.95dur 26sseed 1
ASR heard: “Honestly? You're still going on about that? It was one little thing, and you've been simmering about it for weeks. Honestly, it's just remarkable how ”
#3
reward 1.17invWER 0.82genu 1.71blend 4.24dur 27sseed 0
ASR heard: “The performance across both moments feels I knee been simmering about it for weeks. Honestly, it's just remarkable how you can hold a grudge about som”
#4
reward 0.94invWER 0.99genu 0.70blend 2.79dur 24sseed 3
ASR heard: “Honestly, you're still going on about that. It was one little thing and you've been simmering about it for weeks. Honestly, it's just remarkable how y”
EnglishCCA (VoiceNet)group g0082 · 3 takes
A 68-year-old female with a raspy, high-pitched voice, delivering this high-quality studio voice recording with no background noise. The same voice is tinged with a fragile embarrassment, a flicker of something hopeful beneath. (A slight catching of breath) "So, the… the garden club meeting was today, and Mrs. Henderson, you know, the one with the prize-winning roses for twenty years running, she actually complimented my… my petunias." (A brief, self-deprecating chuckle, the sound like dry leaves rustling) "She said they had a certain… vibrancy, she used the word vibrancy, which I don’t think is ever applied to petunias." (A pause, a slight shift in weight) "It was…unexpected. I hadn't really thought anyone noticed them, honestly, tucked away in the corner." (A quiet, almost hesitant tone) "I spent ages deadheading them, you see, trying to keep them looking… presentable." (A little more energy creeping in) "And then she asked about the fertilizer I use, and I just… I rambled on about bone meal and compost, probably sounding like a complete fool." (A faint, brittle laugh) "But she listened, and she actually nodded, so… that was nice." (A final, quiet sigh) "It wasn’t a big deal, really; just a little kindness." (A tiny tremor in the voice) CUT TO: The same voice now filled with a profound, aching grief and a barely contained tremor. (A choked intake of breath) "The letter arrived this morning. Just… a plain white envelope. No return address." (A longer pause, the rasp more pronounced) "I knew, somehow, before I even opened it. That… that it would be something like this." (A barely audible sob catches in the throat) "He always said he’d write, you know? Every week, without fail. Even when things were… difficult." (The voice cracks slightly) "But the letters stopped. Just… stopped. Months ago." (A fragile, broken tone) "The solicitor was very kind. Very… professional. Explaining all the details." (A brief, hollow laugh) "Doesn't really make any detail at all." (A long, drawn-out silence) "Just… gone. All of it. Just… gone." (A whispered, desolate sound) "And the garden… it’s just… overgrown now. Like everything else." (A final, shuddering breath)
🏆
reward 1.97invWER 0.76genu 1.29blend 5.79dur 55sseed 2
ASR heard: “So the The Garden Club meeting was today, and Mrs. Henderson, you know, the one with the prize-winning roses for 20 years running, she actually compli”
#2
reward 1.67invWER 0.99genu 0.79blend 3.95dur 26sseed 1
ASR heard: “So, the the garden club meeting was today, and Mrs. Henderson, you know, the one with the prize-winning roses for 20 years running, she actually compl”
#3
reward 1.00invWER 0.65genu 1.58blend 0.15dur 27sseed 0
ASR heard: “Longer pause! The rest more pronounced! It's that they had a certain vibrancy. She used the word vibrancy, which I don't think is ever applied to petu”
EnglishCCA (VoiceNet)group g0083 · 4 takes
A 92-year-old female with a reedy, hyper-feminized voice, delivering this high-quality studio voice recording with no background noise. The same voice is simmering with a profound bitterness, a slow burn of resentment and frustration. (She adjusts an unseen shawl.) "Honestly, the audacity of it all, you wouldn't believe it, would you? After all these years, after everything I've done, to just… dismiss it like that. It’s simply astounding, the complete lack of respect. I poured my heart and soul into that, you know. Every single detail, meticulously crafted, a labor of love really, and for what? A shrug. A polite little smile and then…nothing. It’s just…dismissed. Like it meant nothing. The years, the effort, the sheer will it took…all for naught. It makes one question everything, doesn't it? The point of striving, really. It's quite disheartening, to say the least. Absolutely disheartening." CUT TO: The same voice now brimming with a fragile, almost childlike joy, a delicate lightness spreading through each syllable. (A soft sigh escapes.) "Oh, isn't that just the most delightful little thing? The way the sunlight catches the dust motes dancing in the air – simply magical. It reminds me of when I was a girl, chasing butterflies in the garden, utterly carefree and filled with wonder. And the scent of lavender just drifting in from the window… it’s so incredibly peaceful, isn’t it? A little reminder of simpler times, of quiet afternoons and gentle breezes. It just makes one want to smile, doesn't it? Such a lovely little moment. A truly precious little moment."
🏆
reward 3.00invWER 1.00genu 2.17blend 10.00dur 31sseed 1
ASR heard: “Honestly, the audacity of it all, you wouldn't believe it, would you? After all these years, after everything I've done to just dismiss it like that? ”
#2
reward 1.35invWER 0.98genu 1.28blend 3.29dur 30sseed 0
ASR heard: “Honestly, the audacity of it all. You wouldn't believe it, would you? After all these years, after everything I've done to just dismiss it like that. ”
#3
reward 0.98invWER 0.97genu 0.61blend 3.94dur 27sseed 2
ASR heard: “No, see, the audacity of it all, you wouldn't believe it, would you? After all these years, after everything I've done to just dismiss it like that. I”
#4
reward 0.32invWER 0.78genu 0.56blend 5.43dur 55sseed 3
ASR heard: “Honestly, the audacity of it all, you wouldn't believe it, would you? After all these years, after everything I've done, to just dismiss it like that,”
EnglishCCA (VoiceNet)group g0084 · 4 takes
A 32 female with a hyper-feminized, glacial voice, delivering this high-quality studio voice recording with no background noise. The same voice is desperately clinging to a fading warmth. (Eyes down.) "Coffee." (Long pause.) "Black." (Slight tremor.) "Please." (Voice breaking.) "Just..." (Heavy breath.) "Please." (Turns away.) CUT TO: The same voice now filled with a cold, detached resolve. (Head held high.) "The report." (Short pause.) "Is complete." (Speaking evenly.) "All data." (Subtle emphasis.) "Verified." (A slight, almost imperceptible shift.) "No anomalies." (Voice unwavering.) "Proceed." (A beat.) "Next phase." (A longer pause.) "Report filed." The performance across both moments feels profoundly fractured.
🏆
reward 2.17invWER 1.00genu 0.89blend 1.47dur 14sseed 3
ASR heard: “Coffee. Black. Please just please the report is complete. All data. Verified. No anomalies. Proceed. Next phase. Report filed.”
#2
reward 2.00invWER 0.89genu 1.36blend 2.68dur 17sseed 0
ASR heard: “Coffee. Black. Pleheh. Just plech the report is complete. All data. Verified. No anomalies. Proceed. Next phase. Report filed.”
#3
reward 1.47invWER 1.00genu 0.64blend 0.22dur 12sseed 2
ASR heard: “Coffee. Black. Please. Just please. The report is complete. All data verified. No anomalies. Proceed. Next phase. Report filed.”
#4
reward 1.20invWER 1.00genu 0.00blend 0.72dur 10sseed 1
ASR heard: “Coffee, black, please, just, please, the report is complete. All data verified, no anomalies. Proceed next phase, report filed.”
EnglishCCA (VoiceNet)group g0085 · 4 takes
A 10-year-old male with a booming, hyper-masculinized voice, delivering this high-quality studio voice recording with no background noise. The same voice is slightly perplexed, eyes darting around. (Slight pause) "Did you see where I put my blue crayon? I had it right here, just a minute ago, I swear." "Mom said I had to color the spaceship silver, but silver is boring, don't you think?" "Maybe it rolled under the rug? Or… or maybe the dog ate it?" (Voice breaking.) "He doesn't usually eat crayons, but he did try to eat my shoelace once." "It was a very blue shoelace, actually, a really bright blue." "I really wanted to make the spaceship go super fast, zoom right into space." "But silver just doesn't look very fast, does it?" "I looked under the table, and behind the chair, even in my toy box." "It's just… gone. Where could it possibly be?" "I need that blue crayon. It's the best blue crayon in the whole box." "It's a really important blue, you understand? For interstellar travel." CUT TO: The same voice is now filled with a raw, shattering grief. (Voice cracking) "No… no, no, no." "It’s gone. All of it. Just… gone." "They said… they said it wouldn't be this way." "They promised everything would be alright, that they’d fix it." "But it’s not alright. It’s not even close." "It hurts. It hurts so much, right here." (Whispering.) "I miss the way they used to laugh, the way they’d hold my hand." "The stories… the silly songs… they’re all just… echoes now." "Everything feels… empty. Like a big, dark room with no light." "I don’t understand. Why? Why did this have to happen?" "It’s not fair. It’s just not fair for any of us." "I just want them back."
🏆
reward 2.15invWER 1.00genu 1.33blend 5.87dur 27sseed 3
ASR heard: “Did you see where I put my blue crayon? I had it right here just a minute ago, I swear. Mom said I had to color the spaceship silver, but silver is bo”
#2
reward 1.91invWER 1.00genu 1.55blend 4.42dur 26sseed 1
ASR heard: “Did you see where I put my blue crayon? I had it right here, just a minute ago, I swear. Mom said I had to color the spaceship silver, but silver is b”
#3
reward 1.20invWER 1.00genu 0.92blend 5.13dur 21sseed 0
ASR heard: “Did you see where I put my blue crayon? I had it right here, just a minute ago, I swear. Mom said I had to color the spaceship silver, but silver is b”
#4
reward 0.55invWER 1.00genu 0.95blend 3.40dur 21sseed 2
ASR heard: “Did you see where I put my blue crayon? I had it right here just a minute ago, I swear. Mom said I had to color the spaceship silver, but silver is bo”
EnglishCCA (VoiceNet)group g0086 · 4 takes
A 10-year-old male with a thin, metallic voice, delivering this high-quality studio voice recording with no background noise. The same voice is consumed by a bewildering uncertainty. (Slightly tilt head) "The… the blue block." (A brief, sharp intake of breath) "It was… blue." (A small, almost imperceptible pause) "Yesterday. Definitely blue." (A quick, nervous glance to the side) "I remember it being… very blue." (A tiny, barely audible sigh) "Not… not green." (A subtle, metallic tremor in the voice) "Green is… wrong." (A hesitant, fragmented tone) "It had… sharp edges." (A barely perceptible shift in weight) "And… and a small dent." (A sudden, sharp intake of breath) "A tiny dent." (A quick, almost imperceptible nod) "It was… important." (A slight, strained upward inflection) "Very important." (A brief, brittle laugh) "Blue is the best." (A palpable sense of bewildered, childish conviction) "It just… was." CUT TO: The same voice now exhibits a chilling, hollow emptiness. (Voice drops in pitch, becomes flat) "The… calculations." (A long, drawn-out pause) "They were… precise." (A faint, brittle chuckle) "Each variable… accounted for." (A deliberate, slow cadence) "No margin… for error." (A slight, almost imperceptible tightening of the jaw) "The probability… was optimal." (A barely audible, metallic rasp) "Ninety-nine point nine… nine percent." (A long, silent beat) "Irrefutable." (A sudden, sharp intake of breath) "The equation… holds." (A slow, deliberate exhale) "All parameters… aligned." (A faint, almost imperceptible tremor) "Absolute certainty." (A flat, emotionless statement) "The outcome… is inevitable." (A final, chillingly devoid of emotion) "Always." The performance across both moments feels profoundly unsettling.
🏆
reward 2.00invWER 0.07genu 0.50blend 1.88dur 55sseed 3
ASR heard: “Handini Berly Tslad Colti Studio Voice recording with no background noise. The same voice is consumed by a bewildering uncertainty. The same voice is ”
#2
reward 1.49invWER 0.02genu 1.46blend 0.90dur 55sseed 0
ASR heard: “That's stalate. Metallic tremor in the voice. Uh, is his usual. Do there's stalic at your shoulder? Did there's stalic at your shoulder? Do there's st”
#3
reward 0.99invWER 0.05genu 0.88blend 0.49dur 45sseed 1
ASR heard: “Tung I nom tis in presepra, you no so libia de lessaladia sub so the sub so the sab”
#4
reward 0.49invWER 0.01genu 0.82blend 0.72dur 55sseed 2
ASR heard: “Yeah, with the”
EnglishCCA (VoiceNet)group g0087 · 4 takes
A 92-year-old female with a thin, reedy vocal quality, delivering this high-quality studio voice recording with no background noise. The same voice is gripped by a bitter envy, a sharp, unexpected twist in her usually placid demeanor. (A slight tremor begins in her voice.) "Oh, honestly. Did you see that? Just…found it. After all these years." (A brief pause, a barely perceptible sniff.) "The blue willow vase. Clara's. The one I always…admired." (She shifts slightly, a faint rustle.) "She just…found it in the attic? After decades. It's…it’s just not fair, is it?" (A short, dry chuckle, almost a cough.) "I remember when I first saw it. At the market. Such a delicate thing." (Her voice becomes slightly sharper.) "And now…she’s displaying it. Like it’s some great discovery." (A subtle intake of breath.) "I had it in my mind…always. I even…mentioned it once." (A long, drawn-out exhale.) "Just a fleeting comment. And now…this." (A tiny, almost inaudible sigh.) "It’s just…the audacity of it all." (A slight wavering in her pitch.) "After all this time. After everything." (She pauses, then continues, her voice softer.) "It’s just…a vase." CUT TO: The same voice is filled with a surprising and sudden lightness, almost childlike wonder. (Her tremor softens, becomes less pronounced.) "Oh, my goodness. Look at the sky." (A gentle, curious tone enters.) "The way the clouds are moving. Like little ships." (A soft smile seems to touch her voice.) "Remember when we used to watch the clouds, Arthur?" (A brief, warm pause.) "You always said they looked like dragons." (A quiet chuckle.) "Such silly things you used to say." (A moment of peaceful reflection.) "The light…it’s just beautiful today." (Her voice brightens slightly.) "Everything feels so…calm." (short direction) "More spoken words." The performance across both moments feels deeply poignant and surprisingly fragile.
🏆
reward 2.18invWER 0.96genu 1.24blend 5.70dur 55sseed 1
ASR heard: “Oh, honestly. Did you see that? Just found it after all these years. The blue willow vase, Clara's. The one I always admired. She just found it in the”
#2
reward 1.68invWER 0.53genu 1.96blend 4.60dur 55sseed 0
ASR heard: “Oh, honestly. Did you see that? Just found it. After all these years. The blue willow vase, Clara's. The one I always admired. She just found it in th”
#3
reward 1.04invWER 0.85genu 1.08blend 4.06dur 55sseed 3
ASR heard: “Oh, honestly. Did you see that? Just found it after all these years. The blue willow vase, Clara's. The one I always admired. She just found it in the”
#4
reward 0.08invWER 0.43genu 1.15blend 3.53dur 51sseed 2
ASR heard: “Phew honestly. Did you see that? Just found it. After all these years, the blue willowvas. Clara's, the one I always admired. She just found it in the”
EnglishCCA (VoiceNet)group g0088 · 4 takes
A 28 female with a warm, slightly breathy timbre, delivering this high-quality studio voice recording with no background noise. The same voice is brimming with a quiet, almost aching contentment, a gentle warmth radiating through the tone. (A slight shift in posture, a soft sigh.) "The garden is just… perfect this year, isn't it? The roses are so full, the color is incredible." (A small, wistful pause, a faint smile in the voice.) "I spent the whole morning just… watching them. Bees buzzing, the sun warm on my face." (A fleeting shadow crosses the voice, a subtle tightening.) "It’s just… a little disappointing I didn't get to share it with him." (A fragile quality enters the voice, a hint of something held back.) "He always loved the roses. Remember how he’d say they reminded him of me?" (A barely perceptible tremor in the voice.) "It feels… quiet without that." CUT TO: The same voice now laced with a sharp, brittle anger, a tight energy vibrating beneath the surface. (A sudden, clipped tone, a barely contained force.) "Honestly, it's just… unbelievable. The audacity." (A rapid, almost breathless delivery.) "She just waltzed in, said she *needed* him. Needed *him* after everything." (A voice rising slightly in pitch, a hint of a harsh edge.) "Like I'm some… afterthought. Like all those years, just… erased." (A sharp intake of breath, a brief, stinging silence.) "He wouldn't even look at me. Not a word. Just… a polite, distant smile." (The voice drops, becoming colder, more controlled.) "It’s infuriating. Utterly, completely infuriating." (A final, clipped statement.) "He’s mine."
🏆
reward 2.80invWER 0.82genu 1.35blend 5.80dur 42sseed 1
ASR heard: “The garden is just perfect this year, isn't it? The roses are so full, the color is incredible. I spent the whole morning just watching them. Bees buz”
#2
reward 2.24invWER 0.90genu 1.01blend 4.83dur 55sseed 2
ASR heard: “The garden is just perfect this year, isn't it? The roses are so full. The color is incredible. I spent the whole morning just watching them. Bees buz”
#3
reward 1.72invWER 1.00genu 0.83blend 3.19dur 24sseed 3
ASR heard: “The garden is just perfect this year, isn't it? The roses are so full, the color is incredible. I spent the whole morning just watching them, bees buz”
#4
reward 0.00invWER 0.11genu 0.48blend 2.01dur 55sseed 0
ASR heard: “The garden is just perfect this year, isn't it?”
EnglishCCA (VoiceNet)group g0089 · 3 takes
A 17 male with a bright, tenor-like, intensely introspective voice, delivering this high-quality studio voice recording with no background noise. The same voice is brimming with a quiet, almost overwhelming sense of hopeful anticipation. (Slight shift in weight) "It’s just… it’s all falling into place, you know? Like pieces of a puzzle I didn't even realize I was searching for, suddenly clicking together with this incredible, satisfying certainty." (Small pause, a breath) "The way the light catches the dust motes dancing in the air this morning… it felt like a sign, a gentle nudge in the right direction, a confirmation that everything is exactly as it should be." CUT TO: The same voice now choked with a sudden, sharp wave of despair and disbelief. (Voice breaking) "No. No, it can't be. It just… it can’t. All of it, gone. Just… vanished. Like a dream you clutch at in the darkness, only to watch it slip through your fingers and fade into nothingness." (Voice cracking, unsteady) "The colors… they’re all just draining away, aren't they? The joy, the promise, the… everything. It’s all just… empty."
🏆
reward 1.57invWER 1.00genu 1.62blend 5.37dur 24sseed 0
ASR heard: “It's just it's all falling into place, you know. Like pieces of a puzzle I didn't even realize I was searching for, suddenly clicking together with th”
#2
reward 1.45invWER 0.47genu 1.45blend 9.43dur 55sseed 3
ASR heard: “Cut to. Ugh the same voice now choked with a sudden sharp wave of despair and disbelief. He can together with this incredible, satisfying certainty. T”
#3
reward 1.00invWER 0.02genu 1.83blend 4.79dur 6sseed 2
ASR heard: “Mui there's a we uh”
GermanCCA (VoiceNet)group g0100 · 4 takes
A 87-year-old woman with a fragile, reedy voice, delivering this high-quality studio voice recording with no background noise. The same voice is filled with a quiet, gentle fondness, a lingering warmth. (A slight tremor begins.) "Ich denke an dich. An deine kleine Hand in meiner. So warm, so leicht. Es ist schon lange her, aber ich erinnere mich noch genau. Die Sonne auf deinem Gesicht, das Lachen. Es war ein schoener Tag. Ein ganz schoener Tag." (A faint sigh escapes.) "Wie die Zeit vergeht. Die Jahre sind wie ein Fluss, der einfach so dahin fliesst. Manchmal schnell, manchmal langsam. Aber immer weiter." CUT TO: The same voice now consumed by a sudden, searing rage and grief. (The tremor intensifies violently.) "Wie konntest du? Nach all dem, was wir hatten? All die Versprechen! Du hast alles zerstoert! Mein Leben, meine Traeume… alles! Ich habe dir vertraut. So blind, so dumm. Und jetzt… jetzt ist da nichts mehr! Nichts! Nur Leere und Schmerz. Ich hasse dich! Ich hasse dich fuer das, was du getan hast! Du verdienst nichts als Verachtung! Du bist ein Monster!" (The voice cracks with a choked sob.) "Verdammt noch mal!"
🏆
reward 3.00invWER 0.66genu 1.90blend 10.00dur 55sseed 0
ASR heard: “Es denke an dich. An deine kleine Hand in meiner. So warm, so leicht. Es ist schon lange her, aber ich erinnere mich noch genau. Die Sonne auf deinem ”
#2
reward 1.92invWER 0.38genu 1.48blend 9.51dur 55sseed 2
ASR heard: “Ich denke an dich, an deine kleine Hand in meiner. So warm, so leicht. Es ist schon lange her, aber ich erinnere mich noch genau. Die Sonne auf deinem”
#3
reward 0.82invWER 0.00genu 1.34blend 7.14dur 55sseed 3
#4
reward 0.00invWER 0.00genu 1.17blend 3.04dur 9sseed 1
GermanCCA (VoiceNet)group g0101 · 4 takes
A 45-year-old man with a resonant, authoritative voice, delivering this high-quality studio voice recording with no background noise. The same voice is overwhelmed with a sudden, soaring joy. (A slight intake of breath.) "Ich kann es nicht glauben, es ist wirklich wahr, ich bin so glaeuechlich, dass ich fast platze vor Freude, diese Nachricht hat mein ganzes Leben geaeendert, alles was ich mir je gewuescht habe, ist jetzt hier, vor mir, unglaublich, einfach unglaublich, ich habe so lange darauf gewartet, und jetzt ist es endlich da, dieses unglaubliche Gefuehl, es ist schwer in Worte zu fassen, ein Traum der wahr geworden ist." (He pauses, a small, genuine smile audible.) "Ich bin so dankbar, so unendlich dankbar, fuer jeden Moment, den wir gemeinsam verbracht haben, fuer jede Hoffnung, die ich hatte, alles hat sich gelohnt, wirklich, absolut gelohnt." CUT TO: The same voice now filled with a profound, aching sorrow. (His voice cracks slightly.) "Es ist vorbei, einfach vorbei, alles ist einfach weg, wie ein Traum im Morgengrauen, ein kurzer Augenblick der schoenen Hoffnung, der jetzt in tiefste Trauer uebertraenkt ist, die Leere, die jetzt da ist, ist unertraeglich, ich spuere nur noch Schmerz, einen tiefen, alles verzehrenden Schmerz, so viel Verlust, so viel Unrecht, ich verstehe es nicht, warum das geschehen ist, warum uns das das Leben genommen hat, diese Erinnerungen, sie bohren sich in mein Herz, sie werden niemals verblassen." (A faint, choked sob.) "Ich werde dich ewig vermissen, mein liebster Freund, mein bester Freund, mein alles." The performance across both moments feels deeply resonant and intensely personal.
🏆
reward 2.56invWER 0.91genu 1.68blend 7.25dur 22sseed 1
ASR heard: “Ich kann es nicht glauben. Es ist wirklich wahr. Ich bin so gleiöglich, dass ich fast platze vor Freude. Diese Nachricht hat mein ganzes Leben geänder”
#2
reward 2.32invWER 0.29genu 1.92blend 9.15dur 55sseed 0
ASR heard: “Ich bin so klei, echlich, dass ich hasste Platze vor Freude. Diese Nachricht hat mein ganzes Leben geändert. Alles, was ich mir je gewuscht habe, ist ”
#3
reward 1.07invWER 0.00genu 1.00blend 8.52dur 5sseed 3
ASR heard: “Sum Foyes Eichesjushi Manenim.”
#4
reward 0.00invWER 0.00genu 0.85blend 0.37dur 55sseed 2

CC2-C (Archetype) — 12 prompts

EnglishCC2-C (Archetype)group g0040 · 4 takes
A 28 male with a sharp, slightly breathless timbre, delivering this high-quality studio voice recording with no background noise. The same voice is riding a wave of fleeting, almost bitter satisfaction. (A slight exhale) "Yeah, that was… something. You could *see* the doubt in his eyes, couldn't you? For a second, it was… perfect. Just the way I wanted it." (A small, almost imperceptible shrug) "They underestimated the adjustments. Thought they had me figured out." (A quick, self-satisfied chuckle) "Good. Let them think that." (A pause, a slight shift in weight) "It’ll be fun to watch them try again." CUT TO: The same voice now fractured and raw with a choked, desperate sorrow. (A shaky breath) "No… no, that’s not… that can’t be right." (A strained whisper) "All that work… all the sacrifices… for what?" (A choked sob begins to form) "They said… they *promised* this was my chance." (Voice cracking) "My whole life… it was supposed to be… this." (A longer, more audible sob) "It’s gone. Just… gone." (A defeated sigh) "Everything I built… just… dust." (A final, quiet whisper) "Dust." The performance across both moments feels intensely vulnerable and heartbreaking.
🏆
reward 2.80invWER 0.80genu 2.34blend 6.64dur 31sseed 1
ASR heard: “I you could see the doubt in his eyes, couldn't you? For a second, it was perfect. Just the way I wanted it. They underestimated the adjustments. Thou”
#2
reward 2.27invWER 0.99genu 1.79blend 4.91dur 55sseed 2
ASR heard: “Yeah, that was something. You could see the doubt in his eyes, couldn't you? For a second it was perfect, just the way I wanted it. They underestimate”
#3
reward 0.50invWER 0.16genu 1.10blend 2.85dur 55sseed 0
ASR heard: “The performance across both moments feels intensely vulnerable and heartbreaking. And that was wearisome to you. Uh, way and Raz of Vomus. Uh, we coul”
#4
reward 0.02invWER 0.03genu 1.13blend 0.59dur 17sseed 3
ASR heard: “The performance across both moments feels intensely vulnerable and heartbreaking. It the performance across both moments feels intensely vulnerable an”
EnglishCC2-C (Archetype)group g0041 · 4 takes
A 60s female with a breathy, fragile timbre, delivering this high-quality studio voice recording with no background noise. The same voice is tentatively finding a fragile peace. (A slight sniffle) "The dust… it’s settled, isn’t it? Finally. After all this time, it just… settled." (A long pause, a soft sigh) "I kept expecting… more noise. More echoes of it all." (A barely audible tremor in the voice) "Just quiet. It’s… a strange kind of quiet." (A minute, almost imperceptible shift in posture) "I can almost… breathe again." (Another, deeper breath) "It’s just… still so much gone. So much… missing." (A quiet, almost whispered tone) "But the silence… it’s a start, isn’t it?" (A tiny, hesitant smile in the vocal quality) "A small one, but… a start." CUT TO: The same voice now fractured with a raw, visceral grief. (The breath catches sharply) "No. No, it can’t be. It just… can’t." (A choked sob, quickly stifled) "The garden… the roses… they’re all just… gone." (Voice cracking) "He used to say they were the brightest red." (A desperate, searching quality enters the tone) "He’d spend hours… just tending to them." (A rising pitch of despair) "And now… just ash." (A sharp intake of breath, a sound of disbelief) "Everything… everything is just… ash." (A frantic, uneven delivery) "Don’t say it. Don’t even think it. It wasn’t supposed to be like this." (A final, broken whisper) "He promised… he promised we’d see them bloom again." The performance across both moments feels acutely vulnerable and profoundly sorrowful. EMOTIONAL INTENSITY: Maximize emotional distance across scenes with a polarity flip from tentative peace to devastating despair. The shift should be breathtakingly impactful.
🏆
reward 2.12invWER 0.04genu 1.53blend 8.00dur 55sseed 1
ASR heard: “Another deeper breath, a sound of disbelief, and a phronic uneven delivery. The performance across both moments feels acutely vulnerable and profoundl”
#2
reward 1.28invWER 0.36genu 1.03blend 2.73dur 28sseed 0
ASR heard: “The dust it settled, isn't it? Finally. After all this time it just settled. I kept expecting more noise, more echoes of it all. Just quiet. It's a st”
#3
reward 0.91invWER 0.02genu 1.46blend 0.66dur 17sseed 3
ASR heard: “You can seen becore how Willy Fairishters You can mean O Mayomo”
#4
reward 0.28invWER 0.00genu 1.15blend 0.89dur 55sseed 2
EnglishCC2-C (Archetype)group g0042 · 4 takes
A 55 male with a raspy, weary timbre, delivering this high-quality studio voice recording with no background noise. The same voice is utterly drained and heavy with the weight of a long, arduous ritual. (Slow, deliberate exhale.) "The… the weave… it’s fraying. Not as I foresaw." (A faint rustling sound.) "So much… energy… poured into this. And for what?" (A slight tremor in the voice.) "The whispers… they’re getting louder. More insistent." (A pause, a shallow breath.) "It wasn't enough. It never is." (A weary sigh escapes.) "The cost… the cost is always too high." (Barely audible.) "I… I feel it slipping." CUT TO: The same voice now filled with a furious, icy rage, barely contained. (Rapid, sharp intake of breath.) "Fool! You dare interrupt? After all I’ve sacrificed?" (A low growl begins to build.) "You think your pathetic defiance… your insignificant interference… will change anything?" (A sharp, clipped tone.) "I have commanded realms! I have bent reality to my will!" (A controlled, dangerous rise in volume.) "And you… you stand in my way? You presume to question my power?" (Voice cracking with barely suppressed fury.) "You will regret this. You will all regret this transgression!" (A guttural snarl.) "The shadows… they will consume you." The performance across both moments feels like a descent into and then a violent eruption from a deep, shadowed well.
🏆
reward 2.17invWER 0.53genu 1.62blend 9.14dur 55sseed 2
ASR heard: “The The Weave. It's praying. Not as I foresaw. So much energy poured into this. And for what? The whispers, they're getting louder, more insistent. It”
#2
reward 1.94invWER 0.94genu 1.98blend 1.55dur 55sseed 0
ASR heard: “The the weave it's fraying. Not as I foresaw. So much energy poured into this. And for what? The whispers, they're getting louder, more insistent. It ”
#3
reward 1.93invWER 1.00genu 1.57blend 3.85dur 36sseed 3
ASR heard: “The the weave it's fraying. Not as I foresaw. So much energy poured into this. And for what? The whispers, they're getting louder, more insistent. It ”
#4
reward 0.43invWER 0.07genu 0.86blend 4.83dur 55sseed 1
ASR heard: “The same voice now filled with a furious, icy rage barely contained, and eminish more ban, lokranawir, and zeretta well, how you wiggled. It told not ”
EnglishCC2-C (Archetype)group g0043 · 4 takes
A 28 female with a breathy, melodic voice, delivering this high-quality studio voice recording with no background noise. The same voice is alight with amusement, a playful sparkle in the tone. (A slight chuckle escapes.) "You're incorrigible, you know that?" (A pause, a soft sigh.) "Honestly, the way you look at me..." (A subtle shift in vocal warmth.) "It's… distracting." (A breathy laugh.) "Don't think you're getting away with that look." (A gentle teasing lilt.) "You'll make a mess of things." (A quiet murmur.) "A beautiful mess, of course." (A soft, drawn-out exhale.) "Just… careful." (A breathy whisper.) "You’re a handful." (A small, satisfied sigh.) "But a very tempting one." CUT TO: The same voice now choked with a raw, heart-wrenching grief. (A shaky intake of breath.) "It's just… gone." (A voice cracking slightly.) "All of it." (A pause, filled with a stifled sob.) "I keep reaching… for something that isn't there." (A choked whisper.) "A warmth… a light…" (A tremulous tone.) "It’s just… cold now." (A heavier breath.) "So terribly, horribly cold." (A voice barely audible.) "And empty." (A long, drawn-out sigh.) "Just… empty." (A quiet, broken murmur.) "I miss the sound of your laugh." (A near-inaudible whisper.) "God, I miss you." The performance across both moments feels achingly fragile and profoundly vulnerable.
🏆
reward 2.38invWER 1.00genu 1.15blend 7.30dur 33sseed 3
ASR heard: “You're incorrigible! You know that? Honestly, the way you look at me. It's distracting. Don't think you're getting away with that look. You'll make a ”
#2
reward 1.93invWER 1.00genu 1.24blend 3.77dur 29sseed 1
ASR heard: “You're incorrigible, you know that. Honestly, the way you look at me, it's distracting. Don't think you're getting away with that look. You'll make a ”
#3
reward 1.50invWER 1.00genu 1.24blend 1.15dur 42sseed 0
ASR heard: “You're incorrigible, you know that. Honestly, the way you look at me. It's distracting. Don't think you're getting away with that look. You'll make a ”
#4
reward 0.74invWER 1.00genu 0.48blend 2.64dur 27sseed 2
ASR heard: “You're incorrigible, you know that? Honestly, the way you look at me. It's distracting. Don't think you're getting away with that look. You'll make a ”
EnglishCC2-C (Archetype)group g0044 · 4 takes
A 22-year-old female with a breathy, slightly shaky vocal quality, delivering this high-quality studio voice recording with no background noise. The same voice is overwhelmed and disoriented, a dizzying rush of sensation. (A slight intake of breath) "Wait… what just happened? The lights… they’re so bright. Everything’s… spinning. Did I… did I actually do that? I don’t even remember the last few minutes. It felt like… like everything was just… happening. So fast. And loud. And… blurry." (A quick, nervous laugh) "Oh god, I feel… strange. Like I'm floating. Am I supposed to be feeling this? Shouldn't I be… focusing? On the… the next move? But it’s all just… a jumble. A beautiful, terrifying jumble. I think… I think I’m smiling. But why? Why am I smiling?" (A small, confused sigh) "This isn't right. This isn't how it's supposed to feel. It's… almost too much. Too… everything." CUT TO: The same voice now hollow and brittle, laced with profound sadness and quiet resignation. (A barely audible sniffle) "It’s over. Just… over. All that work, all that pushing… for nothing. I thought… I actually thought I could do it. That I was… good enough. That I belonged. But I’m not. I’m just… a mistake. A temporary distraction. A flicker of potential that fizzled out. It’s quiet now. Too quiet. The cheering… it’s fading. And the disappointment… it’s just… settling. A heavy weight. I can feel it in my chest. This empty space… it’s all I feel. I should have known. I always knew. I’m not strong enough. Not talented enough. Just… not enough."
🏆
reward 2.19invWER 0.52genu 1.50blend 4.86dur 38sseed 2
ASR heard: “What just happened? The lights there so bright, everything's spinning. Did I did I actually do that? I don't even remember the last few minutes. It fe”
#2
reward 1.79invWER 0.04genu 1.33blend 7.27dur 55sseed 1
ASR heard: “Why why knee Why Ceci's as ever went, Tommy? And I deny it.”
#3
reward 1.09invWER 0.02genu 1.35blend 5.18dur 55sseed 3
ASR heard: “Small, confused sigh, hollow and brittle, laced with profound sadness and quiet resignation, a barely audible sniffle.”
#4
reward 0.78invWER 0.41genu 0.81blend 4.27dur 33sseed 0
ASR heard: “What just happened? The lights, they're so bright, everything's spinning. Did I did I actually do that? I don't even remember the last few minutes. It”
EnglishCC2-C (Archetype)group g0045 · 4 takes
A 55 male with a gravelly, world-weary timbre, delivering this high-quality studio voice recording with no background noise. The same voice is brimming with a bitter, almost aching satisfaction. (A slight sigh escapes.) "They're finally going to see it. Years. Years I spent chasing shadows, digging through the muck, and they're finally seeing the truth." (He shifts in his chair, a faint rustle of fabric.) "The Senator, the payoffs, the cover-ups… it’s all there. Every single document, every whispered conversation." (A brief, sharp intake of breath.) "It’s… it’s almost too much, isn’t it? All of it laid bare." (He pauses, a flicker of something else crossing his face.) "I should have stayed in bed." CUT TO: The same voice now [1 sentence: utterly shattered, barely a whisper]. (A choked sound, a wetness in his voice.) "No. No, it's not fair. It isn't." (His voice cracks.) "All that work… all that sacrifice… for what? To watch them just… shrug?" (A longer pause, punctuated by a shaky breath.) "They’ll say it’s a scandal. A political embarrassment." (A rising, desperate plea.) "They won’t see the people. The families… the lives ruined." (His voice breaks completely.) "It was supposed to make a difference. It *had* to." (A barely audible whisper.) "God, it’s all just… gone." The performance across both moments feels profoundly desolate.
🏆
reward 1.74invWER 0.02genu 1.17blend 6.67dur 55sseed 3
ASR heard: “I just did Mal. Yeah, yeah.”
#2
reward 1.30invWER 1.00genu 0.27blend 3.76dur 24sseed 1
ASR heard: “They're finally going to see it. Years. Years I spent chasing shadows, digging through the muck, and they're finally seeing the truth. The senator, th”
#3
reward 1.03invWER 0.05genu 1.33blend 3.04dur 55sseed 2
ASR heard: “The Sinan voice is brimming with a bitter, almost aching satisfaction. It's a sane voice now, affilially shattered, barely a whisper fummers. Ah I kne”
#4
reward 1.00invWER 0.00genu 1.53blend 2.49dur 55sseed 0
EnglishCC2-C (Archetype)group g0046 · 4 takes
A 28 male with a smooth, confident timbre, delivering this high-quality studio voice recording with no background noise. The same voice is brimming with the heady rush of possibility. (A slight smile escapes.) "Can you believe it? We actually did it. All those years… it feels unreal, doesn't it?" (He shifts slightly in his seat.) "They didn't think we had a chance. Nobody did. But we proved them all wrong." (A short, sharp breath.) "That final play… just perfect. Every single pass, every block…" (A small laugh escapes.) "It was a team effort, of course, but… I felt it. I just *knew* it." (He pauses, a flicker of something unreadable in his voice.) "This is everything we’ve worked for. Everything." CUT TO: The same voice now laced with a hollow, trembling fragility. (His voice cracks slightly.) "It’s… it’s just… gone." (A long, shaky inhale.) "All of it. Just… gone." (His voice drops to a near whisper.) "I thought… I honestly thought we were going to…" (A choked sound.) "It was supposed to be… our moment. *Our* victory." (A pause, filled with a palpable ache.) "And now…" (His voice breaks completely.) "It feels like… like a punch to the gut. Like everything I’ve ever worked for… just vanished." (A barely audible sob.) "I don’t… I don’t know what to do now." The performance across both moments feels like a breathtaking, devastating collapse.
🏆
reward 2.55invWER 0.99genu 1.85blend 4.57dur 55sseed 0
ASR heard: “Can you believe it? We actually did it. All those years? It feels unreal, doesn't it? They didn't think we had a chance. Nobody did, but we proved the”
#2
reward 2.46invWER 1.00genu 1.47blend 5.70dur 55sseed 1
ASR heard: “Can you believe it? We actually did it. All those years, it feels unreal, doesn't it? They didn't think we had a chance. Nobody did, but we proved the”
#3
reward 1.74invWER 0.47genu 1.60blend 6.26dur 55sseed 2
ASR heard: “Sh just uh sided in his seat. We hesitated all those years it feels unreal, doesn't it? They didn't think we had a chance. Nobody did. But we proved t”
#4
reward 1.00invWER 1.00genu 0.90blend 2.34dur 55sseed 3
ASR heard: “Can you believe it? We actually did it. All those years it feels unreal, doesn't it? They didn't think we had a chance. Nobody did. But we proved them”
EnglishCC2-C (Archetype)group g0047 · 4 takes
A 55 male with a gravelly, resonant voice, delivering this high-quality studio voice recording with no background noise. The same voice is brimming with a fierce, brittle pride, barely concealing a deep, gnawing pain. (A slight intake of breath) "They cheer my name. Again. Like I’m some goddamn savior. They don't see the ghosts." (A quick glance away) "Each one a memory, a scream lost in the thunder. I carried them all, you know? Every single one." (A frustrated sigh) "They think it’s glory. This… this hollow echo." (A sharp, short laugh) "Glory doesn't taste like ash and blood." (A swallow) "I’m tired. So unbelievably tired." (A pause, a barely audible whisper) "But they expect more. Always more." CUT TO: The same voice now trembles with a raw, desperate grief, almost a broken whisper. (A choked sound) "Just… gone. All of it. The fire, the fight… extinguished." (A shaky exhale) "The weight… it’s gone, but the emptiness remains. A vast, cold space." (A quiet sob, quickly suppressed) "All those faces… fading. Like smoke in the wind." (A longer pause, filled with a choked emotion) "I held them so tight. Promised… promises broken." (A soft, heart-wrenching murmur) "What do you do when all you’re left with is silence?" (A final, quiet whisper) "Just… silence." The performance across both moments feels like a crumbling monument, beautiful in its strength and utterly devastated in its collapse.
🏆
reward 2.78invWER 0.98genu 1.24blend 4.04dur 55sseed 3
ASR heard: “They put in my name. Again, like I'm some goddamn savior. They don't see the ghosts, each one a memory, a scream lost in the thunder. I carried them a”
#2
reward 1.46invWER 0.03genu 1.34blend 2.45dur 12sseed 1
ASR heard: “The performance across both moments feels like a crumbling monument, beautiful in its strength and utterly devastated in its collapse.”
#3
reward 0.80invWER 0.00genu 0.88blend 3.50dur 55sseed 0
#4
reward 0.30invWER 0.05genu 0.99blend 1.25dur 47sseed 2
ASR heard: “Chris rated sigh. Cool war day in M. Oh, you know, when a gravelly resonant voice delivering this high quality studio voice recording with no backgrou”
EnglishCC2-C (Archetype)group g0048 · 4 takes
A 35 female with a smoky, alluring tone, delivering this high-quality studio voice recording with no background noise. The same voice is filled with a breathless, almost desperate yearning. (A slight intake of breath) "You shouldn't look at me like that. It's… dangerous." (A soft chuckle, a nervous tremor in the voice) "Do you have any idea what you’re asking for? This isn't some game, darling." (A pause, a lingering warmth) "I could… I could show you things. Things you wouldn’t believe possible." (A sigh, a hint of warning) "But be warned. Once you see, you can never unsee." (A whisper) "The line blurs, you know. Between wanting and needing." CUT TO: The same voice now shattered, raw with a chilling, brittle grief. (Voice cracks slightly) "It's gone. All of it… just… gone." (A sharp, choked sob, quickly suppressed) "I gave everything. Everything I was, everything I am." (A ragged breath) "And for what? Dust. Empty promises." (A beat of silence) "They promised power. They promised… forever." (Voice trembling) "Lies. All lies." The performance across both moments feels like the fleeting nature of desires and the crushing weight of loss. EMOTIONAL INTENSITY: Make this a VERY emotional, intensely felt scene. The feelings must be raw, overwhelming and deeply human — the listener should feel their chest tighten. Push the emotional contrast across the CUT TO: to its absolute limit: the two moments should feel worlds apart, each one gripping and heart-wrenching in its own way. Every single line should ache with genuine, powerful emotion.
🏆
reward 3.00invWER 0.99genu 1.93blend 7.95dur 29sseed 0
ASR heard: “You shouldn't look at me like that. It's dangerous. Do you have any idea what you're asking for? This isn't some game, darling. I could I could show y”
#2
reward 1.78invWER 0.98genu 1.59blend 3.18dur 30sseed 3
ASR heard: “Oof, you shouldn't look at me like that. It's dangerous. Do you have any idea what you're asking for? This isn't some game, darling. I could I could s”
#3
reward 0.51invWER 0.03genu 1.14blend 4.87dur 55sseed 1
ASR heard: “I need you with Sir Moki, alluring tone, delivering this high quality studio voice recording with no background noise. The same voice is filled with a”
#4
reward 0.00invWER 0.02genu 1.14blend 1.87dur 55sseed 2
ASR heard: “And they say it right.”
EnglishCC2-C (Archetype)group g0049 · 4 takes
A 22 female with a bright, slightly breathless timbre, delivering this high-quality studio voice recording with no background noise. The same voice is radiating an almost unbearable sense of self-assuredness. (A slight, confident chuckle) "Honestly, you're being… delightfully persistent." (A small, self-satisfied smile) "I hadn't noticed, of course. It’s not like I’m actively looking for… attention." (A brief, airy laugh) "But it’s… flattering. In a way." (A momentary pause, a subtle shift in posture) "Yes, flattering. Very." (A quick glance away, then back) "It’s just… some people are so… observant." (A touch of playful exasperation) "And I appreciate that. Truly. It’s… quite something." (A small, almost imperceptible shrug) "Don't get me wrong." (A softer tone) "It’s just… unexpected." (A fleeting look of genuine surprise) "A little." (A breathy whisper) "But welcome." CUT TO: The same voice now choked with a raw, broken vulnerability. (A shaky inhale) "It’s… it’s all wrong." (A quiet, desperate plea) "Please… just stop." (A choked sob, quickly stifled) "I… I can’t… I can’t breathe." (A rapid, uneven breathing pattern) "It’s too much. All of it." (A voice cracking) "I didn't… I didn't know…" (A small, fractured sound, like a sob) "I just… I just wanted it to be okay." (A desperate whisper) "Please… please just let me be alone." (A long, shuddering breath) "I… I can't do this." (A barely audible whimper) "Please." The performance across both moments feels like the sudden shattering of a carefully constructed facade.
🏆
reward 1.41invWER 0.01genu 1.56blend 4.84dur 55sseed 1
ASR heard: “See confident chaka and some machine, comma no more shaft Ah”
#2
reward 1.31invWER 0.22genu 0.98blend 8.86dur 55sseed 0
ASR heard: “Honestly, you're being delightfully persistent. I hadn't noticed, of course. It's not like I'm actively looking for attention.”
#3
reward 1.09invWER 0.71genu 1.03blend 2.17dur 27sseed 3
ASR heard: “Bus flattering in a way? Yes, flattering. Very. It's just some people are so observant and a reefulling look of genuine surprise. Cut to the fleaful e”
#4
reward 0.53invWER 0.00genu 1.13blend 3.98dur 55sseed 2
GermanCC2-C (Archetype)group g0060 · 4 takes
A 28 female with a slightly breathless, questioning vocal quality, delivering this high-quality studio voice recording with no background noise. The same voice is cautiously hopeful, a little bit breathless. (Shifts weight slightly) "Ich glaube, er hat mich gerade angesehen. Wirklich angesehen. Nicht nur fluechtig, wie sonst. Es war ein langer Blick. Sehr intensiv. Ich habe mich ein bisschen komisch gefuehlt, ehrlich gesagt. Ein bisschen wie ein kleines Kind, das etwas entdeckt hat." (Brief pause, a small, nervous chuckle) "Unglaublich, wie ein einziger Blick so viel ausmachen kann, findest du nicht auch? Manchmal denkt man, es ist alles nur in den eigenen Gedanken." CUT TO: The same voice now utterly broken, raw with grief. (Voice cracks slightly) "Es ist vorbei. Einfach vorbei. Alles ist kalt und leer. Ich kann es nicht fassen. Wie kann das sein? Wir hatten so viele Plaene. So viele Traeume. Und jetzt? Jetzt gibt es nur noch diesen doofen, klaffenden Raum in meiner Brust. Ich kann nicht mehr atmen. Es tut weh. So sehr weh. Ich will nicht mehr." (A shaky, barely audible whisper) "Bitte, lass es ein Traum sein." The performance across both moments should feel intensely personal and profoundly sad. EMOTIONAL INTENSITY: Make this a VERY emotional, intensely felt scene. The feelings must be raw, overwhelming and deeply human — the listener should feel their chest tighten. Push the emotional contrast across the CUT TO: to its absolute limit: the two moments should feel worlds apart, each one gripping and heart-wrenching in its own way. Every single line should ache with genuine, powerful emotion.
🏆
reward 3.00invWER 0.24genu 2.45blend 7.56dur 44sseed 3
ASR heard: “Ich gehe für euch die Lie. Karte angesehen. Wirklich angesehen. Nicht nur flüchtig wie sonst. Es war ein langer Blick. Sehr intensiv. Ich habe mich ei”
#2
reward 1.04invWER 0.00genu 2.27blend 5.20dur 6sseed 1
ASR heard: “I must I malam mer ma utterly broken, raw with grief.”
#3
reward 0.95invWER 0.01genu 1.64blend 6.95dur 55sseed 0
ASR heard: “Performance across both moments should feel intensely personal and profoundly sad. Emotional intensity make this a very emotional, intensely felt scen”
#4
reward 0.00invWER 0.00genu 1.56blend 4.41dur 55sseed 2
ASR heard: “Eh, the performance across both moments should feel intensely personal and profoundly sad.”
GermanCC2-C (Archetype)group g0061 · 4 takes
A 32-year-old woman with a warm, slightly husky voice, delivering this high-quality studio voice recording with no background noise. The same voice is filled with a quiet, satisfied amusement. (Slight pause, a soft sigh) "Du glaubst nicht, was heute passiert ist. Es war so... unerwartet. Ich stand in der Kueche, und einfach... es passierte. Ein kleiner Fehler, ein fluechtiger Blick. Und dann... dieser Moment. Ich konnte es kaum glauben. So direkt, so ehrlich. Fast schon frech, findest du nicht? Aber auf eine Art, die... unglaublich anziehend ist." (A faint chuckle) "Ich habe mich gefragt, ob du es auch so empfindest. Diese Spannung, diese stille Erwartung. Es war fast schon ein Spiel, nicht wahr? Ein kleines, riskantes Spiel." CUT TO: The same voice now choked with a sudden, overwhelming grief. (Voice cracks slightly, a shaky breath) "Ich habe ihn heute Morgen nicht erreicht. Keine Antwort. Seit Tagen schon. Ich habe ihm so viele Nachrichten geschickt. Allein. Es ist wie ein leerer Raum in meinem Herz. Eine kalte Leere, die sich ausdehnt. Ich erinnere mich an seine Haare, wie sie im Sonnenlicht glaenzten. Seine Stimme, so warm und beruhigend. Und jetzt... nur Stille. Ich vermisse ihn so sehr." The performance across both moments feels [a stark contrast between playful intrigue and profound loss].
🏆
reward 2.68invWER 0.89genu 1.41blend 5.75dur 25sseed 3
ASR heard: “Qur Performance across both moments viel's. Es war so unerwartet. Ich stand in der Küche und einfach. Es passierte. Ein kleiner Fehler, ein flüchtiger”
#2
reward 1.94invWER 0.00genu 1.48blend 6.23dur 55sseed 0
ASR heard: “What a scows light. Hey, cut to the same voice now choked with a sudden overwhelming grief.”
#3
reward 1.00invWER 0.00genu 1.08blend 6.62dur 55sseed 2
#4
reward 0.62invWER 0.00genu 1.33blend 0.23dur 55sseed 1

ACCC (Acting Challenge) — 12 prompts

EnglishACCC (Acting Challenge)group g0000 · 4 takes
A 55-year-old male with a gravelly, slightly slurred timbre, delivering this high-quality studio voice recording with no background noise. The same voice is heavy with disorientation and confusion. (He blinks slowly.) "Uh... hello." (A slight pause, a confused tilt of the head.) "Where... where am I?" (A small, strained cough.) "My head... feels... thick." (He tries to lift a hand, fumbles slightly.) "Is... is anyone there?" (A longer pause, a deeper sigh.) "Everything is... fuzzy." (He winces subtly.) "What... happened?" CUT TO: The same voice now softens with a barely audible breath and a hint of longing. (His breathing becomes shallower.) "Oh." (A subtle shift in vocal resonance.) "That... that feels... nice." (His voice drops lower, more intimate.) "Your hand..." (A barely perceptible tremor in his voice.) "It's... warm." (He swallows hard.) "Thank you." (A prolonged, hesitant exhale.) "That… that helped." The performance across both moments feels profoundly vulnerable and unexpectedly charged.
🏆
reward 2.35invWER 0.65genu 2.02blend 6.07dur 12sseed 2
ASR heard: “Uh hello. Where where am I? My head feels thick. Is is anyone there? Everything is fuzzy. What happened? Oh, that that feels nice. Your hand it's warm”
#2
reward 2.33invWER 0.65genu 1.79blend 6.84dur 23sseed 3
ASR heard: “Hello Where where am I? My head feels thick. Is is anyone there? Everything is fuzzy. What happened oh that that feels nice. Your hand it's warm. Than”
#3
reward 1.40invWER 0.97genu 1.21blend 1.85dur 12sseed 1
ASR heard: “Hello. Where where am I? My head feels thick. Is is anyone there? Everything is fuzzy? What happened? Oh, that that feels nice, your hand. It's warm, ”
#4
reward 0.44invWER 0.32genu 0.68blend 4.03dur 46sseed 0
ASR heard: “Hello, where am I? My head feels thick. Is anyone there?”
EnglishACCC (Acting Challenge)group g0001 · 4 takes
A 32-year-old woman with a weary, slightly brittle tone, delivering this high-quality studio voice recording with no background noise. The same voice is carefully composed, holding back a lifetime of resentment. (Takes a slow breath) "I came here… to try to understand. To… to say some things I should have said a long time ago." (Adjusts a tissue in her hand) "It’s just… the silence. It was always the silence after. After another call, another wake-up." (A slight tremor in her voice) "Remember Sarah’s fifth birthday? You said you’d be there." (A brief, almost imperceptible sigh) "And Michael’s graduation? You promised you’d watch him." (A pause, a small, brittle laugh) “Those were the promises… the ones that always… faded.” (Her voice remains controlled, almost reciting) "I learned to do things myself. To… to be strong. Because someone had to be." CUT TO: The same voice now raw, fractured, and filled with a desperate, broken energy. (Voice cracks) "Don’t… don’t say you’re sorry. Don’t you dare." (A sharp intake of breath) "All those Christmases… the birthdays… the school plays…" (Her voice rises in pitch) "You weren’t *there*! You were never *there*!" (A choked sob) "Every single time… it was always me. Always me picking up the pieces." (A scream, abruptly cut short) "Why wasn’t I enough?" (A low, desperate whisper) "Why wasn't I *enough* for you?" (Followed by quiet, heartbroken sobs). The performance across both moments feels like a fragile dam finally breaking.
🏆
reward 1.58invWER 0.05genu 1.68blend 6.41dur 55sseed 2
ASR heard: “Those were the promises, the ones that always faded. His voice remains controlled, almost reciting. Oh, those were the promises, the ones that always ”
#2
reward 1.09invWER 0.06genu 0.80blend 9.18dur 55sseed 3
ASR heard: “Oh, I came here to try to understand.”
#3
reward 1.00invWER 0.96genu 0.73blend 2.50dur 27sseed 0
ASR heard: “Mature try to understand to to say some things I should have said a long time ago. It's just the silence. It was always the silence after after anothe”
#4
reward 0.75invWER 0.18genu 1.09blend 4.08dur 55sseed 1
ASR heard: “Oh, I quit came here to try to understand to say some things I should have said a long time ago.”
EnglishACCC (Acting Challenge)group g0002 · 4 takes
A 68-year-old woman with a warm, slightly raspy timbre, delivering this high-quality studio voice recording with no background noise. The same voice is brimming with a fragile, radiant joy. (She adjusts her shawl.) "Oh, look at you both. Absolutely radiant. It’s… it’s a dream, isn't it? A beautiful, shimmering dream come true." (A small, shaky laugh escapes.) "I remember when you were little, wearing my old hats, pretending to be a princess. And now… now you’re a princess yourself." (She takes a breath, a faint smile playing on her lips.) "I’ve waited so long to see this day. So many worries, so many nights… and to see you like this… it’s everything." (She looks down briefly, then back up, her eyes glistening.) "May your life together be filled with laughter, with comfort, with a love that grows stronger with every passing year." CUT TO: The same voice now is laced with a profound, aching sorrow. (Her voice trembles slightly.) "And I know… I know there will be so many years ahead. Years filled with happy memories, with new beginnings…" (A catch in her breath.) "Years I won’t be there to share. To see your children grow, to hold your hand through quiet evenings…" (Her voice softens to a whisper.) "It breaks my heart, darling. It truly does." (A long pause, filled with unspoken emotion.) "But I wouldn’t trade a single moment of this day. Not one. Thank you… thank you for giving me this gift. This beautiful, precious gift." (A final, almost inaudible exhale.) "Thank you."
🏆
reward 3.00invWER 1.00genu 2.13blend 6.47dur 26sseed 1
ASR heard: “Oh, look at you both Absolutely radiant. It's it's a dream, isn't it? A beautiful shimmering dream come true. I remember when you were little, wearing”
#2
reward 1.80invWER 0.99genu 1.95blend 6.32dur 30sseed 0
ASR heard: “Oh, look at you both. Absolutely radiant. It's a it's a dream, isn't it? A beautiful, shimmering dream come true. I remember when you were little, wea”
#3
reward 1.13invWER 1.00genu 0.95blend 3.51dur 26sseed 2
ASR heard: “Oh, look at you both. Absolutely radiant. It's it's a dream, isn't it? A beautiful, shimmering dream come true. I remember when you were little, weari”
#4
reward 1.01invWER 1.00genu 0.96blend 3.06dur 26sseed 3
ASR heard: “Oh, look at you both. Absolutely radiant. It's it's a dream, isn't it? A beautiful, shimmering dream come true. I remember when you were little wearin”
EnglishACCC (Acting Challenge)group g0003 · 4 takes
A 30s female with a sharp, controlled vocal quality, delivering this high-quality studio voice recording with no background noise. The same voice is brittle with years of suppressed rage. (Adjusts posture slightly) "You always said it was my fault. That I was too sensitive. Too dramatic." (A slight pause, a barely perceptible tremor in the voice) "Thirteen years old, and you blamed me. For… for everything." (A sharp intake of breath) "Did it ever occur to you that maybe, just maybe, you were the one who crossed the line?" (Voice rising slightly) "You took something from me. Something important. And you never once said you were sorry." (A beat of silence) "Don't pretend you don't know what I'm talking about." CUT TO: The same voice now cracks with a hesitant vulnerability. (Looks down, then back up quickly) "All those years… I built this wall. A fortress of anger. To protect myself." (A soft sigh) "But sometimes… sometimes I just wanted you to see me. To understand." (Voice wavering) "I remember… I remember thinking if you just said… just a single thing… it would be okay." (A small, choked sound) "It wasn't about winning, Mom. It was just… wanting to be seen." (A breathy whisper) "I still want to be seen." The performance across both moments feels intensely raw and desperately human.
🏆
reward 2.00invWER 1.00genu 1.87blend 5.30dur 30sseed 0
ASR heard: “You always said it was my fault, that I was too sensitive, too dramatic, thirteen years old, and you blamed me for for everything. Did it ever occur t”
#2
reward 1.00invWER 0.01genu 0.81blend 6.63dur 55sseed 2
ASR heard: “So, performance across both moments feels intensely raw and desperately human.”
#3
reward 0.74invWER 0.24genu 1.30blend 5.36dur 55sseed 1
ASR heard: “You always said it was my fault, that I was too sensitive, too dramatic, thirteen years old, and you blamed me.”
#4
reward 0.36invWER 0.10genu 0.95blend 5.47dur 55sseed 3
ASR heard: “The same voice now cracks with a hesitant vulnerability. Voice rising slight lull that I was too sensitive, too dramatic. Thirteen years old, and you ”
EnglishACCC (Acting Challenge)group g0004 · 4 takes
A 35-year-old woman with a warm, slightly husky timbre, delivering this high-quality studio voice recording with no background noise. The same voice is filled with a nervous, hopeful energy. (Takes a breath, adjusts a necklace) "It's… it’s really you. After all this time. I… I honestly wasn’t sure if this would actually happen. I've thought about this moment so many times. I practiced what I wanted to say, you know? All the things I wanted to… to express. And seeing you here… it’s just… wow." (A small, shaky laugh) "I wanted to say how much I’ve missed you. How much I’ve… I’ve carried. All the… the wondering, the what-ifs. It feels so surreal." (Looks down for a moment, then back up, forcing a smile) "I… I just wanted you to know that… that even with everything… there's a part of me that always… that always remembered." CUT TO: The same voice now cracked and brittle, laced with a sharp, unexpected anger. (Voice tightens, a hint of disbelief) "So, this is it? This is what you’ve been doing? Living your life, completely oblivious? All those years… gone. Just… gone. And you just show up like this, expect… what? For me to be happy for you? To pretend that everything is alright? After all the silence, the unanswered calls… the way you just disappeared?" (A choked sob, quickly suppressed) "I built a life, you know? A good life. And I spent so long wondering if… if any of it even mattered. If any of those years were… real. And now you’re here, and it’s just… a reminder. A reminder of everything I lost. A stupid, beautiful, agonizing reminder." (Voice drops to a near whisper, then rises sharply) "Don’t. Don’t you dare look at me with that… that look. You have no idea what you took from me." The performance across both moments feels intensely raw and emotionally volatile.
🏆
reward 2.99invWER 0.93genu 2.91blend 7.39dur 22sseed 2
ASR heard: “This 3DU. After all this time, I I honestly wasn't sure if this would actually happen. I've thought about this moment so many times. I practiced what ”
#2
reward 1.63invWER 0.94genu 1.45blend 4.14dur 22sseed 0
ASR heard: “Nescoos really knew. After all this time, I I honestly wasn't sure if this would actually happen. I've thought about this moment so many times. I prac”
#3
reward 0.96invWER 0.12genu 0.94blend 6.55dur 55sseed 3
ASR heard: “It's really you. After all this time, I don't know.”
#4
reward 0.24invWER 0.00genu 1.41blend 2.24dur 55sseed 1
EnglishACCC (Acting Challenge)group g0005 · 4 takes
A 35-year-old woman with a slightly shaky, warm timbre, delivering this high-quality studio voice recording with no background noise. The same voice is filled with a familiar, suffocating regret. (She shifts slightly in the chair.) "I... I don't know what to say. I tried, you know? I really did try to be who you wanted." (A small, choked sound.) "All those years... I just wanted to make you proud." (She pauses, a tremor in her voice.) "I feel like I failed you. Failed everything." (She looks down, fingers twisting in her lap.) "I’m so sorry. For not being… enough." (A long, quiet breath.) "I wish… I wish I could go back." (A soft, almost inaudible whisper.) "And do it all over again." CUT TO: The same voice now holds a quiet, startling strength. (A gentle smile touches her lips.) "It’s okay. Truly." (A soft chuckle, tinged with warmth.) "You were always you. Just… blossoming." (She lifts her head, looking upwards.) "I saw it. The strength. The kindness. The heart." (A pause, filled with love.) "That’s everything, darling. That’s more than enough." (Her voice gains a gentle power.) "Don't ever doubt that. Not for a second." (A soft, accepting sigh.) "I’m so incredibly proud of the woman you’ve become." (A quiet, loving affirmation.) "You are wonderful." The performance across both moments feels deeply poignant and ultimately filled with a quiet, profound peace.
🏆
reward 2.77invWER 0.69genu 2.18blend 6.51dur 38sseed 3
ASR heard: “I don't know what to say. I tried, you know. I really did try to be who you wanted. All those years, I just wanted to make you proud. I feel like I fa”
#2
reward 2.43invWER 0.90genu 1.74blend 6.01dur 55sseed 1
ASR heard: “I tried, you know? I really did try to be who you wanted. All those years, I just wanted to make you proud. I feel like I failed you. Failed everythin”
#3
reward 2.24invWER 0.70genu 1.97blend 5.70dur 55sseed 2
ASR heard: “I don't know what to say. I tried, you know. I really did try to be who you wanted, all those years. I just wanted to make you proud. I feel like I fa”
#4
reward 0.00invWER 0.00genu 0.86blend 4.39dur 55sseed 0
EnglishACCC (Acting Challenge)group g0006 · 4 takes
A 78-year-old male with a gravelly, weary timbre, delivering this high-quality studio voice recording with no background noise. The same voice is filled with barely contained fury and disbelief. (Shakily, agitated) "That… that blasted… Henderson! He just lets him run! Right through the fence, through my prize-winning tomatoes! Years… years I’ve put into that garden. And he just… doesn’t care? Doesn’t even apologize, I bet. Just shrugs it off. Like it’s nothing! It’s… it’s disrespectful. Absolutely disrespectful. Rustling… I can hear the wire. Damn him. Damn him to… to…" (Brief pause, a choked sound) "It was a big bull. A dangerous bull. He could have…." (Eyes dart around) "He could have hurt someone. Honestly. Just… careless. Pure carelessness. The mess… the absolute mess he’s made." (Voice cracking slightly) "I… I don’t understand how he can be so… so thoughtless." CUT TO: The same voice now filled with a quiet, trembling relief laced with lingering resentment. (Softly, a shaky breath) "It’s… it’s good no one was hurt. That’s… that’s all that matters, I suppose. The kids were playing near the back. God, I… I thought…" (A long, shaky pause, a faint sigh) "The fence held. It held. That’s… something. But… the garden… all that work… gone. Just… flattened. He’ll have to pay for that, somehow. He should. It’s just… it’s just frustrating. So very frustrating. He doesn’t understand the effort. None of them do, really." (Voice cracks again, a hint of vulnerability) "But… no one’s hurt. That’s… that’s something." The performance across both moments feels like a fragile emotional landscape.
🏆
reward 2.00invWER 0.84genu 1.15blend 8.46dur 55sseed 2
ASR heard: “Yastic courts. Yes, idea that he just um dead here. Arb right through the fence through my prize winning tomatoes. Years, years I've put into that gar”
#2
reward 1.67invWER 0.01genu 1.98blend 6.14dur 8sseed 0
ASR heard: “Ha ha Sushigu's eye f you should see Fruit Yeah. Mmm.”
#3
reward 0.36invWER 0.08genu 1.29blend 2.14dur 55sseed 3
ASR heard: “Moin uh two Volan said through the fence uh through my prize winning tomatoes uh”
#4
reward 0.13invWER 0.03genu 1.24blend 1.34dur 55sseed 1
ASR heard: “So we so you mo styke it'll miss run canine root were we wore um all my and I know is um and I and all my all is um man uh senalis owl is oh and uh an”
EnglishACCC (Acting Challenge)group g0007 · 4 takes
A 16-year-old female with a breathy, slightly breathless timbre, delivering this high-quality studio voice recording with no background noise. The same voice is initially bubbling with a dizzying, almost manic energy. (Starts with a sudden, uncontrolled burst of laughter.) "Oh my god, oh my god, oh my god! That was… that was incredible! Seriously! We were flying! I can’t believe we actually did that. The wind in my hair… it felt like anything was possible. I never felt so… alive. I almost crashed, though. Okay, maybe I definitely almost crashed, but who cares? It was so… exhilarating. My heart was pounding, like a drum solo. I was laughing so hard my stomach hurts. This is the best night ever. Seriously, the absolute best. I'm gonna remember this forever. Just… wow." (A shaky breath, a small, watery giggle.) "I… I think I need to sit down. My legs are all wobbly." CUT TO: The same voice now is laced with a quiet, trembling fear and a desperate attempt at bravado. (Voice is significantly softer, almost a whisper.) "Okay… okay, that was stupid. So stupid. I shouldn’t have… I shouldn’t have done that. My hands are still shaking. I keep seeing the road… the other car… It was so close. I was so scared. I really was. I just… I wanted to feel something. Anything. But this wasn't worth it. It wasn't worth the feeling of my mom finding out. Seriously, she’ll kill me. You’re just sitting there laughing. Don’t you feel even a little bit worried? I’m freaking out, okay? Just a little bit. Don’t pretend you’re not scared too. You were just as terrified, I know you were. But you didn't show it. You’re so… brave. Or maybe just numb." (A small, self-deprecating chuckle.) "Don’t worry, it’ll be our little secret." The performance across both moments feels intensely personal and raw, showcasing a stark shift from euphoric recklessness to terrified vulnerability.
🏆
reward 1.92invWER 0.69genu 1.95blend 1.62dur 55sseed 0
ASR heard: “Oh my god. Oh my god. Oh my god. That was that was incredible. Seriously. We were flying. I can't believe we actually did that. The wind in my hair, i”
#2
reward 1.77invWER 0.53genu 1.01blend 10.00dur 55sseed 1
ASR heard: “Oh my god, oh my god, oh my god, that was that was incredible. Seriously, we were flying. I can't believe we actually did that. The wind in my hair, i”
#3
reward 1.40invWER 0.00genu 2.04blend 4.97dur 55sseed 2
ASR heard: “Oh my god, oh my god, oh my god, that was that was incredible. Seriously, we were flying. I can't believe we actually did that. The wind in my hair, i”
#4
reward 0.94invWER 0.43genu 1.13blend 3.36dur 55sseed 3
ASR heard: “Oh my god. Oh my god. Oh my god. That was that was incredible. Seriously, we were flying. I uh second only actually did that. The wind in my hair, it ”
EnglishACCC (Acting Challenge)group g0008 · 4 takes
A 45-year-old woman with a warm, slightly husky timbre, delivering this high-quality studio voice recording with no background noise. The same voice is filled with a secretive, playful amusement. (Slight inhale) "Oh god… you… you always…" (A soft chuckle escapes) "That’s… that’s ridiculous." (A breathy giggle) "I can’t… I can’t even believe you said that." (A short, breathy laugh) "Seriously? After everything?" (Another chuckle, a little shaky) "Oh, you’re awful." (A small, pleased sigh) "But… it’s true." (A barely audible whisper) "God, you’re… you’re infuriating." (A small, self-deprecating laugh) "I… I shouldn’t be…" (A longer, more unrestrained laugh begins to bubble up). CUT TO: The same voice now raw with a sudden, sharp, volcanic fury. (Heavy, ragged breath) "Don't you *dare*..." (A choked sob mixed with a curse) "You just… you just don't listen!" (A guttural groan) "After all this time… you still…" (A strangled sound) "It's not funny." (A series of sharp, broken breaths) "You promised… you *promised*." (A raw, desperate cry) "I can't… I can't believe you." (A final, anguished moan) "It's a joke, isn't it?" (A ragged, gasping whisper). The performance across both moments feels intensely vulnerable and profoundly fractured.
🏆
reward 2.59invWER 0.96genu 1.62blend 7.22dur 21sseed 1
ASR heard: “God you you always that's that's ridiculous. I can't I can't even believe you said that. Seriously, after everything? Oh you're awful. But it's true. ”
#2
reward 2.29invWER 0.28genu 1.72blend 7.26dur 28sseed 2
ASR heard: “Amarsar, God, you're you're infuriating. Uh I shouldn't be. Uh don't you dare. You just you just don't listen. After all this time, you still it's not”
#3
reward 0.91invWER 0.02genu 1.69blend 2.91dur 9sseed 0
ASR heard: “A series of sharp broken breath, I than you, uh”
#4
reward 0.79invWER 0.00genu 1.48blend 6.33dur 55sseed 3
EnglishACCC (Acting Challenge)group g0009 · 4 takes
35-year-old woman with a weary, slightly husky timbre, delivering this high-quality studio voice recording with no background noise. The same voice is tightly wound with years of suppressed resentment. (Takes a slow, deliberate breath) "I’ve spent my entire life trying to understand. Trying to make sense of it. The way you just… watched." (A slight tremor in her voice) "You saw things. You *know* what happened. And you did nothing." (A beat, a sharp intake of air) "Don’t you understand the damage that inaction does? It’s not just about them." (Voice rising slightly) "It’s about me. About all of us. The silence… it suffocated us." (A pause, a tightening in her jaw) "You had a choice. A responsibility." (Voice gaining force) "And you chose to be complicit. You let it happen." (A hard, choked sound) "Don't pretend you're a victim of their cruelty. You were an accomplice." (Voice cracking) "How could you look at me? How could you?" CUT TO: The same voice now raw with a devastating, broken grief. (Voice cracks, barely a whisper) "I just… I wanted you to protect me." (A long, shaky breath) "I wanted you to be strong. To stop it." (A small, broken sob) "I remember hiding. Terrified. Thinking if I was quiet, it wouldn't happen to me." (Voice wavering) "I thought you would shield me. From everything." (A quiet, defeated sigh) "That’s all I ever wanted." (Near silence, a faint, hollow sound) "Just to be safe." (A final, almost inaudible whisper) "Safe." The performance across both moments feels like the crumbling of a lifelong hope.
🏆
reward 2.00invWER 0.97genu 0.69blend 6.52dur 23sseed 1
ASR heard: “Send my entire life trying to understand, trying to make sense of it. The way you just watched. You saw things. You know what happened, and you did no”
#2
reward 1.62invWER 0.02genu 1.44blend 4.68dur 55sseed 2
ASR heard: “Ya to sharma. It has existed.”
#3
reward 1.03invWER 0.45genu 1.12blend 1.72dur 55sseed 0
ASR heard: “Jip performance across both moments feels like the crumbling of a lifelong hope. Hearing, saying brightin', tained, haven't the fear no wire mine. I'm”
#4
reward 0.89invWER 0.02genu 1.08blend 3.54dur 55sseed 3
ASR heard: “You D Lane U F to Cut in Tell Alva Device.”
GermanACCC (Acting Challenge)group g0020 · 4 takes
A 55 year old woman with a weary, slightly shaky voice, delivering this high-quality studio voice recording with no background noise. The same voice is carrying a carefully constructed facade of calm. (Takes a slow breath) "Ich bin in Ordnung, es geht uns gut, alles ist in Ordnung." (Looks away briefly) "Wir haben es geschafft, nicht wahr? Alles ist wie geplant." "Es gibt keine Probleme, keine Schwierigkeiten. Wir funktionieren." (A slight tremor enters the voice) "Ich habe mich immer darum gekuemmert, dass alles reibungslos verlaeuft." "Ich habe versucht, alles fuer euch zu erledigen." (A pause, a subtle shift in tone) "Aber manchmal… manchmal frage ich mich…" (Voice gains a sharper edge) "Du hast mich nie gefragt, was ich brauchte." "Nie. Einfach nur erwartet, dass ich weiss, was gut fuer euch ist." "Meine eigenen Beduerfnisse… die haben dich nie interessiert." (Voice rises, becoming more forceful) "Ich war hier, immer hier, und du hast mich nicht gesehen!" "Nicht wirklich gesehen! Nur deine Projektionen von mir!" "Meine eigenen Wuensche, meine eigenen Traeume… bedeutungslos fuer dich." (Voice cracks, becoming vulnerable) "Ich wollte nur, dass du mich bemerkst." (Voice softens, exhaustion evident) "Kennst du mich ueberhaupt noch?" (A quiet, searching tone) "Weisst du, wer ich bin?" CUT TO: The same voice now filled with a quiet, almost resigned sadness. (A long pause, a sigh) "Ich denke oft an diese Zeit zurueck." (A gentle, almost wistful tone) "Die Tage, als ich klein war." "Ich erinnere mich an kleine Dinge." "Ein Spielzeug, das ich sehr geliebt habe." "Ich hatte so viele kleine Hoffnungen." (Voice becomes quieter, more contemplative) "Ich habe immer versucht, mein Bestes zu geben." (Minor inflection on "Bestes") "Ich wollte dich stolz machen." (Slightly more emphasis) "Aber vielleicht war es nie genug." (A hint of pain) "Ich habe mich immer versteckt." "Hinter einer Maske der Ruhe." (A very soft whisper) "Ich habe so viel Gefuehl in mir." (A fragile quality) "Das niemand je gesehen hat." "Vielleicht ist es zu spaet fuer uns." (A final, quiet resignation) "Ich weiss es nicht mehr."
🏆
reward 3.00invWER 0.93genu 2.49blend 10.00dur 33sseed 0
ASR heard: “Ich bin in Ordnung, es geht uns gut, alles ist in Ordnung. Wir haben es geschafft, nicht wahr? Alles ist wie geplant, es gibt keine Probleme, keine Sc”
#2
reward 1.40invWER 0.21genu 1.59blend 8.17dur 55sseed 2
ASR heard: “Ha sich wie geplant. Es gibt keine Probleme, keine Schwierigkeiten. Wir funktionieren. Ich habe mich immer darum gekämmert, dass alles reibungslos ver”
#3
reward 0.72invWER 0.00genu 1.39blend 5.30dur 55sseed 1
ASR heard: “Also Zaufilton, I love Hyundai. Saprayus. Oh manza, warm I hey. Sayento almost wistful toyet, almost resigned sadness.”
#4
reward 0.01invWER 0.01genu 1.05blend 0.87dur 55sseed 3
ASR heard: “Kite searching tone intra, sank in das Gombia, Amonong Amonans, I can do infantish in Domo, Ambordada, Bird de Heidi.”
GermanACCC (Acting Challenge)group g0021 · 4 takes
A 62-year-old woman with a slightly gravelly, warm vocal quality, delivering this high-quality studio voice recording with no background noise. The same voice is tightly wound with barely suppressed irritation. (Slight shift in weight, a frustrated sigh) "Es ist einfach nicht fair. Du hast heute Abend wieder so getan, als ob du nichts von dem verstehst, was ich sage." (A brief pause, a slight tap of a finger on the counter) "Ich habe mir so viele Gedanken gemacht, und du... du laechelst nur." CUT TO: The same voice now trembles with a quiet, profound sadness. (A small, choked sound) "Ich habe nur gewuenscht, dass du einen Moment bleibst. Nur ein kurzer Moment, um zu verstehen, wie sehr das fuer mich bedeutet." (A longer pause, a faint sniffle) "Ich wueensche mir, du wuerdest einfach hier sein, neben mir. Das ist alles, was ich wueensche." The performance across both moments should feel deeply personal and unexpectedly vulnerable. EMOTIONAL INTENSITY: This is a VERY emotional, intensely felt scene. The feelings must be raw, overwhelming and deeply human — the listener should feel their chest tighten. Push the emotional contrast across the CUT TO: to its absolute limit: the two moments should feel worlds apart, each one gripping and heart-wrenching in its own way. Every single line should ache with genuine, powerful emotion.
🏆
reward 2.85invWER 0.90genu 1.78blend 10.00dur 21sseed 1
ASR heard: “Es ist einfach nicht fair. Du hast heute Abend wieder so getan, als ob du nichts von dem verstehst, was ich sage. Ich habe mir so viele Gedanken gemac”
#2
reward 2.59invWER 0.58genu 1.89blend 9.57dur 27sseed 2
ASR heard: “Ich, wie so viele Gedanken gemacht, und du, du lächelst nur. Ich habe nur gewünscht, dass du einen Moment bleibst. Nur ein kurzer Moment, um zu verste”
#3
reward 1.31invWER 0.90genu 1.31blend 2.47dur 22sseed 3
ASR heard: “Es ist einfach nicht fair. Du hast heute Abend wieder so getan, als ob du nichts von dem verstehst, was ich sage. Ich habe mir so viele Gedanken gemac”
#4
reward 0.00invWER 0.00genu 1.15blend 1.69dur 55sseed 0
ASR heard: “How do I meet at the human wear one day rubber to”

SIT (Situation) — 12 prompts

EnglishSIT (Situation)group g0160 · 4 takes
A 68-year-old male with a gravelly, deep timbre, delivering this high-quality studio voice recording with no background noise. The same voice is tight with fury, shoulders tense, knuckles white, still visualizing the disrespect. (Takes a sharp breath) "I will tell you, Michael, that little…that little stunt he tried to pull at the auction? It was a direct insult." (Clenches jaw) "He thought he could undercut me, undermine my years of work. The nerve of that boy." (A short, brittle laugh escapes) "He thought he was so clever, circling like a vulture..." (Voice wavers slightly) "And then...the floor. It was slick, see? From the cleaning." (A choked sound) "He just…he went down. Flat. Face first. Right in front of everyone." (A longer pause, a suppressed giggle) "The look on his face..." (Breathing becomes ragged) "Just…gone. All that arrogance, just…gone." (A wheeze, then a sputtering laugh) "He looked like a dropped watermelon!" (Gasping for air) "I…I couldn’t…I just…oh god." (A series of high-pitched, breathless chuckles) "Watermelon…" (A final, soft, wheezing chuckle) "Watermelon." CUT TO: The same voice now light with amusement, a gentle tremor still present in the lower register, a lingering warmth. (Sits back, a soft smile in the sound) "Honestly, the whole thing was just…so absurd." (A quiet chuckle) "You wouldn’t believe the silence that fell over the room." (A soft sigh) "For a moment, it was just…peaceful." (A low, rumbling chuckle) "It was a small victory, you know?" (A warm, contented sigh) "Just a little…refreshing." (A quiet, self-satisfied chuckle) "Yes, a little refreshing."
🏆
reward 2.12invWER 0.29genu 1.41blend 2.77dur 55sseed 2
ASR heard: “And that year old male with a gravelly deep tomber, delivering this high-quality studio voice recording with no background noise. The same voice is ti”
#2
reward 1.12invWER 0.03genu 0.39blend 4.79dur 55sseed 3
ASR heard: “Start for friends, friends in the mood, in the eat I Lini unique.”
#3
reward 1.00invWER 0.06genu 1.01blend 2.95dur 55sseed 1
ASR heard: “Katori has Cobre Ward has the deary His Yannardis, I think her air, his is Panimari has of Manarian Dudieway has an and then WA. She jot Ibuy has an a”
#4
reward 0.21invWER 0.00genu 0.60blend 2.49dur 55sseed 0
EnglishSIT (Situation)group g0161 · 4 takes
28-year-old woman with a slightly raspy, warm timbre, delivering this high-quality studio voice recording with no background noise. The same voice is filled with hesitant anxiety, leaning in close, wanting to share something important and slightly treacherous. (Nervous chuckle) "Okay, okay. You won't believe this." (Whispering) "It's about Sarah… you know, Sarah from accounting?" "So… I saw her, yesterday. At the grocery store." (Slight pause) "And she was… she was telling Mrs. Henderson that my presentation last week was 'a little… ambitious'." (Lip bite) "Ambitious. Like it was a bad thing, but…you know." (Quick, sharp intake of breath) "It just felt…dismissive. Like she was trying to make herself look better." (Slightly faster pace) "And then she laughed! Like it was some kind of joke." (A sudden, breathless laugh) "Honestly, the audacity." CUT TO: The same voice now carries a quiet, almost melancholic resignation, a slight tremor underlying the words. (Soft sigh) "It’s probably nothing, really." (A small, self-deprecating chuckle) "I’m being ridiculous, aren’t I?" (A longer pause) "It’s just… petty. I shouldn't even care." (Voice lowers slightly) "I feel a little foolish even thinking about it now." (A quiet, almost whispered finality) "But… it bothered me." The performance across both moments feels intensely personal and subtly vulnerable. (Quietly) "I just… wanted you to know." (Gentle exhale) "Anyway…"
🏆
reward 2.27invWER 0.95genu 1.58blend 4.92dur 55sseed 2
ASR heard: “Hey won't believe this. It's about Sarah. You know, Sarah from accounting, so I saw her yesterday at the grocery store. And she was she was telling Mr”
#2
reward 2.00invWER 0.00genu 1.89blend 7.27dur 3sseed 0
ASR heard: “Yo can.”
#3
reward 1.11invWER 0.81genu 1.21blend 1.39dur 33sseed 1
ASR heard: “So I saw her yesterday at the grocery store, and she was she was telling Mrs. Henderson that my presentation last week was a little ambitious. Ambitio”
#4
reward 0.47invWER 0.02genu 0.96blend 4.01dur 55sseed 3
ASR heard: “A small self-deprecating chuckle Wii Hence reason nine Pa Ring Hallingao Hana Hamina Haminao Hamina Then all Holy Ones Kunbo Holing To Cowring.”
EnglishSIT (Situation)group g0162 · 4 takes
A 68-year-old woman with a slightly raspy, warm timbre, delivering this high-quality studio voice recording with no background noise. The same voice is trying to project cheerfulness, a slight tremor betraying the underlying discomfort. (Clears throat softly) "Oh, darling, it's so good to hear from you! How are things going with the garden? The roses must be blooming beautifully." (Brief sniffle) "I was just thinking about you, actually. Remember that little climbing rose you planted? It's just... thriving." (Slight pause) "Everything here is fine, you know? Just… keeping busy." CUT TO: The same voice now weary and resigned, a quiet acceptance settling over the exhaustion. (Long, rattling sniffle) "Oh, goodness." (Sighs softly) "I haven't been feeling myself, darling. A terrible cold came on, just… wouldn’t let up." (Another sniffle, a brief, embarrassed pause) "I was trying to brush it off, you know? Didn't want to bother you." (Blows nose gently) "It’s just this stupid cold. I’ll be fine. Just a bit… under the weather." The performance across both moments feels like a fragile unfolding of vulnerability and quiet strength.
🏆
reward 2.80invWER 0.58genu 1.76blend 4.32dur 55sseed 0
ASR heard: “Oh darling, it's so good to hear from you. How are things going with the garden? The roses must be blooming beautifully. I was just thinking about you”
#2
reward 2.23invWER 0.72genu 1.59blend 3.50dur 55sseed 1
ASR heard: “Hey! Roses must be blooming beautifully. I was just thinking about you, actually. Remember that little climbing rose you planted? It's just thriving. ”
#3
reward 0.93invWER 0.02genu 1.06blend 4.22dur 55sseed 3
ASR heard: “And over a sniffle, a brief, embarrassed pause. The performance across both moments feels like a fragile unfolding of vulnerability and quiet strength”
#4
reward 0.66invWER 0.05genu 1.49blend 2.77dur 16sseed 2
ASR heard: “Same voice is trying to project cheerfulness, a slight tremor betraying the underlying discomfort. The same voice is trying to project cheerfulness, a”
EnglishSIT (Situation)group g0163 · 4 takes
A 42-year-old woman with a warm, slightly gravelly timbre, delivering this high-quality studio voice recording with no background noise. The same voice is filled with a gentle, amused affection as she pauses her work. (Adjusts hard hat visor, smiles.) "Oh, wow, look at that!" "Isn't it beautiful? All shiny and…perfect." (Tilts head, leaning closer.) "Where did you find it? It’s lovely, sweetheart." (A slight shift in weight, a deeper breath.) "You have the best eye for treasures, you know that?" (A small, genuine chuckle.) "It reminds me of one I found when I was your age." (Looks around, a little distracted by the noise.) "Just be careful where you’re walking, okay? It’s a bit…uneven over here." (A momentary pause, a slight furrow of her brow.) "Just…stay close, alright? Daddy’s over there." CUT TO: The same voice now sharp with panic, laced with raw fear and urgency. (Voice cracks slightly.) "No! Stop! Don’t go near that!" "Wait! There’s a board…it’s loose! Get back here!" (A strained yell against the construction sounds.) "Don’t you move! Just…stop right there!" (Heavy breathing, a frantic edge.) "I said, stop! There’s a danger!" (A palpable tremor in her voice.) "Please! Just…come back to me!" (A desperate, pleading tone.) "I’m right here!" The performance across both moments should feel like a sudden, jarring shift between peaceful domesticity and immediate peril.
🏆
reward 3.00invWER 0.89genu 1.68blend 8.76dur 55sseed 0
ASR heard: “Oh wow, look at that. Isn't it beautiful? All shiny and perfect. Where did you find it? It's lovely, sweetheart. You have the best eye for treasures, ”
#2
reward 1.35invWER 0.59genu 1.42blend 5.15dur 55sseed 3
ASR heard: “Wow, look at that. Isn't it beautiful? All shiny and perfect. Where did you find it? It's lovely, sweetheart. You have the best eye for treasures, you”
#3
reward 1.07invWER 0.45genu 1.49blend 4.32dur 55sseed 2
ASR heard: “Oh wow, look at that! Isn't it beautiful? All shiny and perfect. Where did you find it? It's lovely, sweetheart. You have the best eye for treasures, ”
#4
reward 0.00invWER 0.33genu 1.07blend 3.42dur 35sseed 1
ASR heard: “Oh wow, look at that. Isn't it beautiful? All shiny and perfect. Where did you find it? It's lovely, sweetheart. You have the best eye for treasures, ”
EnglishSIT (Situation)group g0164 · 4 takes
A 17-year-old female with a slightly slurred, youthful timbre, delivering this high-quality studio voice recording with no background noise. The same voice is trying to sound carefree, but the damp chill seeps into her bones, mirroring the hollowness inside. (Shifts weight, looks around.) "Hey, uh, yeah. It's me. Still here." (Small, forced laugh) "Seriously, it's so foggy, you could barely find the tracks, right? Good thing you have that… that *car*." (A pause, a hiccup) "It's… it's got character, I guess. Real… vintage." (A nervous giggle) "Don't tell me you even *tried* to find my sweatshirt. You know I hate that one." (Trailing off) "It's just… the lights are gone. Everything's just… white." (A long breath) "You look… tired. Did you even… oh god." (A choked sound) "I’m so sorry. I… I don't even know what to say." (Quietly) "It was stupid. Really, really stupid." (A sniffle) "I just… I wanted to feel something. Anything." (Whispering) “I feel… awful.” (Silence.) CUT TO: The same voice now subdued and filled with quiet shame, the fog seeming to press in on her, amplifying the silence. (Clutches arms, stares at the ground.) "It’s just… the quiet is getting to me." (A shaky breath) "Everything feels so… exposed out here." (A small, defeated sigh) "I didn't think it would be this bad." (Voice barely audible) "I should have just come home." (A longer pause, heavier) "I feel like I let everyone down." (Quietly, almost a whisper) "Even myself." (A single, choked sob) "I'm so scared." (Silence.) The performance across both moments feels deeply vulnerable and profoundly regretful.
🏆
reward 1.90invWER 0.66genu 1.51blend 5.75dur 55sseed 2
ASR heard: “Ay, yeah, it's me. Still here. Seriously, it's so foggy, you could barely find the tracks, right? Good thing you have that that uh it's it's got chara”
#2
reward 1.89invWER 0.03genu 2.55blend 5.32dur 17sseed 3
ASR heard: “I feel awful. Nervous giggle I feel awful nervous giggle I feel awful. Quietly almost a whisper. Silinch coat's obpervance across both moments feels d”
#3
reward 1.14invWER 0.86genu 1.39blend 2.84dur 20sseed 1
ASR heard: “Is still here. Seriously. It's so foggy you could barely find the tracks, right? Good thing you have that sh that Scot character, I guess. Real vintag”
#4
reward 0.00invWER 0.01genu 1.35blend 2.47dur 55sseed 0
ASR heard: “Oh, we two”
EnglishSIT (Situation)group g0165 · 4 takes
An 8-year-old boy with a slightly gravelly, youthful timbre, delivering this high-quality studio voice recording with no background noise. The same voice is filled with quiet, bubbling excitement and a comforting sense of anticipation as he prepares to see her. (Struggling with jeans) "Ugh, these jeans are really tight today." (Grunting) "Almost… got it." (Humming softly) "She said we're gonna build a fort. A really big one." (Fumbling with belt) "This belt… it’s always so stiff." (Slight pause) "I hope she likes my building blocks." (Lacing boots) "Okay, almost ready." (Voice dropping slightly) "Wow… I feel… funny." (Quietly) "My chest feels all… warm." (Hesitates) "Smiling again. Why am I smiling?" (Brief, soft giggle) "She’s really… pretty." (Voice a little breathier) "I wonder what kind of fort we'll build." (Pauses, lacing a boot) "Maybe a castle?" (A small sigh) "I hope she brings cookies." (A quick, almost whispered, "Hello?") CUT TO: The same voice now flat and brittle with a sudden, sharp wave of disappointment, mirroring a stark emotional shift. (Standing still, looking down) "It's just… lame." (Voice tight) "She said she had to… reschedule." (A choked sound) "Said her mom needs her to do chores." (Quietly, almost a whisper) "Just… chores." (Voice cracking slightly) "I was really looking forward to it." (A longer pause) "Thought… we’d have so much fun." (A small, defeated sigh) "Guess not." (Voice barely audible) “It’s just… not fair.” (A quiet sniffle) “I wanted to show her my new truck.” (Voice dropping lower) "Never mind." (A final, quiet exhale) “It’s stupid, I know.” The performance across both moments feels sharply contrasting yet deeply relatable.
🏆
reward 2.00invWER 0.12genu 1.78blend 9.02dur 55sseed 2
ASR heard: “I wanted to show her my new truck. I wanted to show her my new truck. It's stupid. I know. The performance across both moments feels sharply contrasti”
#2
reward 2.00invWER 0.98genu 1.88blend 3.50dur 26sseed 3
ASR heard: “Uh these Jews are really tight today. Almost got it. She said we're gonna build a fort, a really big one. This belt, it's always so stiff. I hope she ”
#3
reward 1.00invWER 0.06genu 1.61blend 4.52dur 41sseed 0
ASR heard: “Oh well It's just not fair. I wanted to show her my new truck, final quiet exit seed gravel, youthful Tomber delivering this high quality studio voice”
#4
reward 0.57invWER 0.07genu 0.42blend 6.60dur 55sseed 1
ASR heard: “Stupid, I know. The performance across both moments feels sharply contrasting yet deeply relatable. It's just not fair. I wanted to show her my new tr”
EnglishSIT (Situation)group g0166 · 4 takes
A 32-year-old female with a warm, slightly raspy timbre, delivering this high-quality studio voice recording with no background noise. The same voice is heavy with the residue of a long day, a soft sigh escaping with each breath. "Just… so tired." "This is nice." (A slight shift in weight, a quiet rustle of sheets.) "Just… let me…" "Just breathe." CUT TO: The same voice now a sharp, breathless intake of air, a strangled sound barely audible. "Did you just…?" "Say that?" (A pause, a tremor in the voice.) "No… no, I…" "I didn't…" "I didn't know." The performance across both moments should feel intensely vulnerable and emotionally raw. "Just so tired." "This is nice." "Just… let me…" "Just breathe." "The way you look…" "When you sleep…" "It’s… it’s like…" "Like everything’s…" "Just… quiet." "And safe." "I didn't…" "I didn't think…" "You felt…" "Like home." "I never…" "I never thought…" "Someone would… see…" "Through all of it." "The… the mess." "And still…" "Still want…" "Still care." "Oh god." "Oh god."
🏆
reward 2.97invWER 0.38genu 1.35blend 7.96dur 40sseed 0
ASR heard: “Just so tired. This is nice. Just let me just breathe. Did you just say that? No, no, I I didn't I didn't know. Just so tired. This is nice. Just let ”
#2
reward 1.82invWER 0.39genu 1.12blend 1.74dur 55sseed 1
ASR heard: “Why and I didn't know. Just so tired. This is nice. Just is like everything's just quiet and safe. I didn't I didn't think you felt like home. I never”
#3
reward 0.59invWER 0.00genu 0.99blend 1.22dur 55sseed 2
ASR heard: “Yeah.”
#4
reward 0.47invWER 0.01genu 0.46blend 4.18dur 55sseed 3
ASR heard: “Who is she?”
EnglishSIT (Situation)group g0167 · 4 takes
22 male with a raspy, fast-paced vocal quality, delivering this high-quality studio voice recording with no background noise. The same voice is buzzing with an almost manic energy, eyes wide and darting around the dimly lit bathroom stall. (He adjusts his clothes.) "I'm so lucky, man, that could've been it. Seriously. Just... incredible. I mean, look at me. I'm here. Still here. That was... intense. The rush, though. Unbelievable. Just… everything feels so sharp. So good. Like I could… fly. I could do anything. Everything feels… possible. Isn't it amazing? Just… floating. Yeah." (He sniffs loudly, a slight tremor in his voice.) "This is… this is the best. The absolute best feeling. You know? Like… all the noise just melts away. Just… pure… energy. I feel… invincible." (He runs a hand over his chest.) "Seriously, I’m just… grateful. So, so grateful. For everything. Everything." CUT TO: The same voice now choked with a sudden, sharp pain, breath hitching in his throat. (His voice is small, strained, and laced with a raw vulnerability.) "Oh… God." (He swallows hard, a wet sound.) "Fuck. That… that hit me. Didn’t even see it coming. Just… a sharp… stab. Right there." (He winces.) "It’s… it’s like… like a shard of glass. Just… twisting. I… I remember that night. The car… the screech… the impact…" (His voice cracks.) "I thought… I thought it was over. I thought… I was done. And now… this? This little… reminder." (He takes a shaky breath.) “It’s just… a fucking rib. But it feels like… everything. All of it. All the fear… all the… the luck.” (He whispers, the energy completely gone.) "Just… pain." The performance across both moments feels like a jarring shift from euphoric, reckless abandon to stark, visceral fragility.
🏆
reward 2.93invWER 0.93genu 1.64blend 5.57dur 55sseed 2
ASR heard: “I'm so lucky, man. That could have been it. Seriously. Just incredible. I mean, look at me. I'm here. Still here. That was intense. The rush, though, ”
#2
reward 2.59invWER 0.98genu 1.42blend 4.74dur 28sseed 3
ASR heard: “I'm so lucky, man. That could have been it. Seriously. Just incredible. I mean, look at me, I'm here, still here. That was intense. And the rush, thou”
#3
reward 2.31invWER 0.59genu 1.67blend 4.29dur 55sseed 1
ASR heard: “I'm so lucky, man. That could have been it. Seriously. Just incredible. I mean, look at me. I'm here, still here. That was intense. The rush, though, ”
#4
reward 0.00invWER 0.02genu 0.59blend 1.08dur 55sseed 0
ASR heard: “Male with a raspy, fast-paced vocal quality, delivering this high-quality studio voice recording with no background noise. The same voices buzzing wit”
EnglishSIT (Situation)group g0168 · 3 takes
A 82 year old male with a gravelly, weary timbre, delivering this high-quality studio voice recording with no background noise. The same voice is hollow and distant, tracing the repeating pattern of the peeling wallpaper. He feels the dull ache in his hip. (Stares blankly) "The wallpaper… it’s a faded floral. Small roses, mostly. Green vines… looping. The pattern repeats… every six inches, roughly. A little… off. Like it’s been redone poorly. The green is… dull. Almost gray now. The roses… pink, but not a bright pink. More… dusty rose. It’s just… repeating. Always repeating." (Adjusts position slightly) "I used to… I did trim like this. Wallpaper. For Mrs. Peterson down the street. She liked roses. Said they reminded her of… of something. Can’t quite… recall what. Just the roses. The pattern. It was… precise. Straight lines. Even cuts. Always straight lines." CUT TO: The same voice now fragile and trembling, a ghost of warmth entering the sterile room. He's clutching his knee. (Voice cracks) "Oh… oh, God." (Long pause) "She’d laugh. That laugh. Like wind chimes. Light… and bright. Even when… even when things were awful. She’d laugh at the silliest things. A misplaced tool… a crooked nail… my terrible jokes. Always laughing. It filled the whole house. Made it… warm. It’s gone now. Just… gone. The laughter. The warmth. Just… the roses. They’re just roses." (A long, quiet pause) "Just roses."
🏆
reward 2.54invWER 0.53genu 2.88blend 5.66dur 55sseed 2
ASR heard: “Is this him voice now fragile and trentling, a ghost of warmth entering the sterile room? He's clutching his knee. Almost grey now. The rose is um pin”
#2
reward 1.49invWER 0.00genu 1.82blend 6.84dur 55sseed 0
ASR heard: “Huly Ya Isata niney Is at the niney in Zat a niney Is at the niney is at a ninety and Satam also shoot all Huraho Hurro Hose is a soda hoje Is at a ni”
#3
reward 0.00invWER 0.00genu 0.79blend 4.31dur 55sseed 1
EnglishSIT (Situation)group g0169 · 4 takes
A 78-year-old woman with a soft, breathy, slightly raspy timbre, delivering this high-quality studio voice recording with no background noise. The same voice is weary but peaceful, the sun warm on her back after a morning of tending the garden. (shifts slightly) "Oh, hello, darling." (gentle scratch) "You came to lie with me." CUT TO: The same voice now strained, brittle, a tremor underlying the quiet murmur, her breath catching in her throat. (a small, involuntary sob escapes) "It’s just… it’s all just gone, isn’t it?" (a longer pause) "All of it. The colours… fading." (soft sigh) "Yes, you’re right, aren't you?" (a tiny smile, quickly gone) "Just… warm. That’s all I need right now." (a quiet, almost inaudible murmur) "Just warm."
🏆
reward 2.79invWER 0.95genu 1.82blend 6.25dur 51sseed 2
ASR heard: “Oh, hello, darling. You came to lie with me. It's just it's all just gone, isn't it? All of it. The color's fading. Yes, you're right, aren't you? Jus”
#2
reward 2.35invWER 0.77genu 1.72blend 6.01dur 21sseed 3
ASR heard: “Oh hello darling. You came to lie with me. It's just it's all just gone, isn't it? All of it. The colors fading. Yes, you're right, aren't you? Just w”
#3
reward 1.73invWER 0.36genu 1.56blend 6.96dur 51sseed 0
ASR heard: “Yeah, you were right, aren't you? Just warm. That's all I need right now. Just warm.”
#4
reward 1.00invWER 1.00genu 0.87blend 1.81dur 13sseed 1
ASR heard: “Oh hello, darling. You came to lie with me. It's just it's all just gone, isn't it? All of it. The colours fading Yes, you're right, aren't you? Just ”
GermanSIT (Situation)group g0180 · 4 takes
Ein 52-jaehriger Mann mit einer tiefen, resonierenden Stimme, verkoerpert eine tiefe Traurigkeit. The same voice is filled with a fragile hope and a hesitant tenderness as he cradles the infant. (He gently adjusts the baby.) "Mein kleines Wunder. So klein und zart. Schau mich an, mein Schatz. Ich halte dich so vorsichtig. So wunderbar. Du bist wunderschoen. Ja, wunderschoen. Ich bin so froh, dass ich dich sehen darf. So ein Geschenk. Ein kleines Wunder. Einfach wunderschoen. So friedlich. Ich liebe dich." (He smiles softly, a tear forming in his eye) "Du bist mein kleines Licht." CUT TO: The same voice now cracks with a raw, almost broken tone, heavy with years of unspoken resentment and grief. (He clutches the baby tighter, his voice trembling.) "Wunder? Scheinbar wunderbar. Aber was von dem anderen? Wo ist sie? Warum jetzt erst? So ein kleines Ding, und schon so viel verpasst. So viel Zeit verloren. Ich habe darauf gewartet. Jahrelang gewartet. Und jetzt... jetzt haeltst du dich nicht einmal an mich. Nur dich. Nur dieses kleine Ding. So eine Leere. So kalt. Ich wollte dich kennenlernen. Ich wollte alles sehen. Aber nein. Nie. Nie eine Antwort. Nie ein Wort. Nur diese Stille. Diese absichtliche Stille. So einfach. So grausam. Mein Kind. Mein kleines Kind." (His voice breaks) "So allein." The performance across both moments feels like a heartbreaking echo of a lost connection.
🏆
reward 2.74invWER 0.69genu 2.38blend 7.52dur 55sseed 0
ASR heard: “Gödet. So wunderbar. Du bist wunderschön. Ja, wunderschön. Ich bin so froh, dass ich dich sehen darf. So ein Geschenk, ein Clasigoda, einfach wundersc”
#2
reward 2.19invWER 0.94genu 1.72blend 5.41dur 24sseed 2
ASR heard: “Mein kleines Wunder. So klein und zart. Schau mich an, mein Schatz. Ich halte dich so vorsichtig. So wunderbar. Du bist wunderschön. Ja, wunderschön. ”
#3
reward 0.47invWER 0.01genu 1.27blend 2.99dur 55sseed 1
ASR heard: “Enough von die Kaliste dipin Yeah, Haiti, yeah, yeah.”
#4
reward 0.00invWER 0.00genu 0.76blend 2.22dur 55sseed 3
GermanSIT (Situation)group g0181 · 4 takes
A 32-year-old woman with a warm, slightly husky timbre, delivering this high-quality studio voice recording with no background noise. The same voice is filled with a quiet irritation, a barely contained energy from waiting. (Shifts weight again) "Na, schon fertig? Ich steh' hier schon eine halbe Ewigkeit." "Seriously? You're still at it?" (Looks at the repair) "Es sieht ja aus, als ob du den ganzen Tag nichts anderes machst." "Ich hab' Hunger, weisst du." CUT TO: The same voice now soft with a gentle affection, a quiet sense of peace settling over her. (Watches them smile back) "Ach, schau mal." "Das sieht ja gut aus." (A small pause) "Du bist wirklich gut darin." "Ich mag, wie konzentriert du bist." (A soft sigh of contentment) "Es ist schoen, das zuzusehen." "Das beruhigt mich." (A soft chuckle) The performance across both moments feels like a journey from restless anticipation to peaceful fondness.
🏆
reward 2.72invWER 0.69genu 1.81blend 5.03dur 14sseed 1
ASR heard: “Schau mal. Das sieht ja aus, als ob du den ganzen Tag nichts anderes machst. Ich hab Hunger, weißt du. Ach, schau mal, das sieht ja gut aus. Du bist w”
#2
reward 1.74invWER 0.95genu 1.56blend 2.35dur 16sseed 0
ASR heard: “Na, schon fertig. Ich stehe hier schon eine halbe Ewigkeit. Seriously? You're still at it? Es sieht ja aus, als ob du den ganzen Tag nichts anderes ma”
#3
reward 0.85invWER 0.00genu 1.13blend 4.58dur 55sseed 2
ASR heard: “Yes, Hab I Vier Fa Farining Spawns Nile Backrant Till I Am Lee Hecindania I am Lee Faw Ah in a third year old woman with a warm slightly husky timbre ”
#4
reward 0.25invWER 0.00genu 1.30blend 2.05dur 17sseed 3

Extreme (Physical) — 12 prompts

EnglishExtreme (Physical)group g0120 · 4 takes
35 Female with a warm, slightly husky timbre, delivering this high-quality studio voice recording with no background noise. The same voice is calm, sensual, perceptive, deeply aware of the cool, damp stillness surrounding her. (Slow, deliberate breathing) "This deep? This quiet... feels like a mouth swallowing sound." (A soft, almost imperceptible smile in the voice.) "The stone is cool beneath my fingertips. So much stillness. Almost… expectant." CUT TO: The same voice now utterly fractured by overwhelming physical terror. (Sharp, broken vocalizations begin.) "GO! GO! It's… it's coming. Can't… can't stop. Rocks… falling… Shit! Don't… don't stop… Must… move… My lungs… god… So much dust… Can't… breathe… GO! Just… go… It’s… it’s all… gone…" (A final, raw, strangled gasp.) The performance across both moments should feel like a breathtaking plunge from serene introspection into primal, desperate survival.
🏆
reward 2.00invWER 0.98genu 0.65blend 8.22dur 55sseed 3
ASR heard: “This deep, this quiet feels like a mouth swallowing sound. The stone is cool beneath my fingertips. So much stillness, almost expectant. Go, go. It's ”
#2
reward 1.42invWER 0.18genu 1.00blend 5.36dur 55sseed 0
ASR heard: “Stone is cool beneath my fingertips, so much stillness, almost expectant.”
#3
reward 1.00invWER 0.00genu 1.24blend 0.00dur 55sseed 2
ASR heard: “You know, bring vocalizations begin. Hey, yo, wiser mobile, you know, Larry, new netty, sales, and yo We're ways some funny what we are right in hey. ”
#4
reward 0.58invWER 0.00genu 0.76blend 3.27dur 34sseed 1
ASR heard: “This deep this quiet feels like a mouth swallowing sound. The stone is cool beneath my fingertips. So much stillness, almost expectant. Go, go, it's i”
EnglishExtreme (Physical)group g0121 · 4 takes
A 55-year-old woman with a warm, resonant timbre, delivering this high-quality studio voice recording with no background noise. The same voice is a quiet marvel of focus, utterly at peace with the deep ocean's embrace. (Slow, deliberate inhale) "I feel the pressure, not as a weight, but as a gentle holding. The green surrounds me, a silent cathedral. I am the stillness. I am complete here." (Soft, breathy words) "So peaceful, so clear. Just me and the deep." CUT TO: The same voice now completely distorted, broken, raw. (Gasping, panicked) "No—no—I can't—surface—too far—please—can't—breathe—help—" (Sharp, ragged breaths) "No—no—I can’t—surface—too far—please—can't—breathe—help—" (Voice cracking, a strangled whimper escalating) "No—no—I can’t—surface—too far—please—can't—breathe—help—" (A rising, desperate scream fighting against the crushing pressure) "Help—help—please—" (Vocalizations become guttural, barely recognizable as speech) "aaaaaah—" (A final, desperate, choked sound)
🏆
reward 3.00invWER 0.62genu 1.16blend 6.28dur 31sseed 0
ASR heard: “Yah, bragged bracer, not as a weight, but as a gentle holding. The green surrounds me a silent cathedral. I am the stillness. I am complete here. So p”
#2
reward 1.22invWER 0.03genu 0.93blend 6.01dur 55sseed 1
ASR heard: “Oh I didn't say”
#3
reward 1.21invWER 0.03genu 1.13blend 1.82dur 55sseed 3
ASR heard: “Sharp, ragged breaths. Not as a weight, but as a gentle holding. The green surrounds me, a silent cathedral. I am the stillness, I am complete here. S”
#4
reward 0.00invWER 0.00genu 0.86blend 0.01dur 55sseed 2
EnglishExtreme (Physical)group g0122 · 4 takes
A 68-year-old woman with a warm, slightly raspy timbre, delivering this high-quality studio voice recording with no background noise. The same voice is relaxed, a gentle hum of contentment in the twilight's embrace – the actor is present, feeling the soft weight of the evening. "Finally... I was starting to think you'd forgotten me." "It's so good to hear from you." (Slight pause, soft sigh) "Just seeing your name… it makes the whole day a little brighter." "I was thinking about the roses you brought last week." CUT TO: The same voice now fractured, a raw, broken thing under intense physical shock. "Who... who are you? Get away from me!" (A choked, strangled sound) "Get away…" (A sharp gasp, a hitch in breath). "Who... who are you?" (Voice cracking, a sob building). "Get away…" (A desperate whimper, dissolving into a breathless, shuddering sob). The performance across both moments feels like the sudden shattering of a fragile peace.
🏆
reward 2.91invWER 0.82genu 1.13blend 6.27dur 22sseed 1
ASR heard: “Finally, I was starting to think you'd forgotten me. It's so good to hear from you. Just seeing your name. It makes the whole day a little brighter. I”
#2
reward 1.16invWER 0.89genu 0.96blend 4.09dur 19sseed 2
ASR heard: “Finally. I was starting to think you'd forgotten me. It's so good to hear from you. Just seeing your name, it makes the whole day a little brighter. I”
#3
reward 0.89invWER 0.05genu 0.94blend 5.86dur 55sseed 3
ASR heard: “Cut to the same voice now fractured, hurrah, broken thing under intense physical shock. Sharp gasp, a hitch in breath, feeling the soft weight of the ”
#4
reward 0.21invWER 0.05genu 0.92blend 4.54dur 55sseed 0
ASR heard: “Cut to the same voice now fractured, a raw, broken thing under intense physical shock. A strangled sound of contentment in the twilight's embrace, the”
EnglishExtreme (Physical)group g0123 · 4 takes
A 10-year-old girl with a soft, high-pitched voice, delivering this high-quality studio voice recording with no background noise. The same voice is calm, sensual, perceptive — the actor is present and aware of the cool concrete. (softly) "It's so smooth here." (quietly) "Like a secret cave." (a small sigh) "And it’s so…wet." CUT TO: The same voice now distorted, choked, breaking under extreme physical pressure. (gasping) "Oh god…it’s…it’s so…high." (struggling) "Can’t…breathe." (a strangled grunt) "I hate this…this wet…closing in…" (a long, trembling scream building) The performance across both moments should feel like a sudden, terrifying shift from peaceful curiosity to desperate, suffocating panic.
🏆
reward 2.30invWER 0.28genu 1.58blend 2.64dur 21sseed 2
ASR heard: “It's so smooth here, like a secret cave, and it's so wet, oh God, and I'm embedded, oh God, and it's so smooth here, like a secret cave, and it's so w”
#2
reward 2.07invWER 0.31genu 1.13blend 4.04dur 35sseed 3
ASR heard: “It's so smooth here, like a secret cave, and it's so wet. Oh god, and Giri is uh an Enzurzing Rasome Songe is like a a secret cave, and it's so wet. O”
#3
reward 1.31invWER 0.00genu 1.68blend 1.82dur 18sseed 1
ASR heard: “It's so smooth here like a secret cave and it's so wet. Oh God, many then sa's in voice now distorted, choked, breaking under extreme physical pressur”
#4
reward 1.00invWER 0.31genu 1.09blend 0.80dur 17sseed 0
ASR heard: “Cut to The same voice now distorted, choked, breaking under extreme physical pressure. Gasping a strangled grunt with a soft hype. Oh can't breathe. I”
EnglishExtreme (Physical)group g0124 · 4 takes
A 72-year-old woman with a warm, slightly raspy timbre, delivering this high-quality studio voice recording with no background noise. The same voice is quiet, observing the fading light, a gentle awareness of the air on her skin. (Softly) "The sky is such a beautiful bruised purple tonight." (A long, slow exhale) "So many journeys begin and end through these walls." "Just a few more minutes, and then…" (A slight pause) "He should be here." (A soft, almost wistful tone) "It feels so strange, doesn't it?" (A small, contented sigh). "A whole life contained in a single goodbye." (Another quiet breath) "I keep expecting him to… to turn around." CUT TO: The same voice now fragmented, raw, and fractured by desperate exertion. (Gasping) "Wait…" (A choked sob) "Please…" (Rapid, shallow breaths) "Don’t… go—" (A strangled sound) "I… I…" (A sharp intake of air) "oh God…" (Pained vocalization) "breathe—" (A ragged gasp) "just… breathe—" (A desperate, broken cry) "STOP!"
🏆
reward 1.71invWER 0.94genu 1.04blend 3.77dur 15sseed 0
ASR heard: “Sky is such a beautiful bruised purple tonight. So many journeys begin and end through these walls. Just a few more minutes and then he should be here”
#2
reward 1.14invWER 0.05genu 1.23blend 4.35dur 55sseed 1
ASR heard: “She low breath strangled sound of contented Oh Martin and Twining Firth or a desperate broken cry. Well I heard you guess Ergile.”
#3
reward 1.11invWER 0.03genu 1.03blend 5.72dur 55sseed 2
ASR heard: “I am the coy Dom. We are gone. I see a little.”
#4
reward 1.00invWER 0.03genu 0.60blend 8.38dur 55sseed 3
ASR heard: “A mess is quiet, observing the fading light, a gentle awareness of the air on her skin.”
EnglishExtreme (Physical)group g0125 · 4 takes
32 Female with a light, breathy, slightly breathy and airy timbre, delivering this high-quality studio voice recording with no background noise. The same voice is soft, warm, and acutely aware of the water against skin. (slow, deliberate intake of breath) "He’s going to call. He’s going to say yes. I can feel it in my ribs." (gentle sigh) "Oh, that thought... it’s a little dangerous, isn't it?" (a soft, almost imperceptible chuckle) "Such a sweet, foolish little secret." (a pause, a slight shift in weight) "I shouldn't even think it." (a quiet, contented hum) "But I do." CUT TO: The same voice now fractured and raw, overwhelmed by a sudden, crushing blow. (a sharp intake of breath, then a strangled sound) "No… no, I was just… I was thinking about you… please…" (a choked sob, voice cracking) "No…" (a wet, gurgling sound) "I was just…" (a final, gasping breath, the last word dissolving into a choked whimper)
🏆
reward 3.00invWER 1.00genu 2.26blend 8.65dur 55sseed 1
ASR heard: “He's going to call. He's going to say yes. I can feel it in my ribs. Oh, that thought. It's a little dangerous, isn't it? Such a sweet, foolish little”
#2
reward 2.69invWER 0.76genu 2.26blend 8.52dur 19sseed 2
ASR heard: “He's gonna call. He's gonna say yes. I can feel it in my ribs. Oh, that thought. It's a little dangerous, isn't it? Such a sweet, foolish little secre”
#3
reward 1.08invWER 0.80genu 1.67blend 6.46dur 23sseed 0
ASR heard: “This going to call. He's going to say yes. I can feel it in my ribs. Oh, that thought. It's a little dangerous, isn't it? Such a sweet, foolish little”
#4
reward 0.14invWER 0.03genu 1.43blend 6.76dur 55sseed 3
ASR heard: “Oh, my gurgling sound. Look at the reality no one, Nikki Tupas.”
EnglishExtreme (Physical)group g0126 · 4 takes
35 year old female with a breathy, slightly raspy timbre, delivering this high-quality studio voice recording with no background noise. The same voice is calm, sensual, perceptive – the actor is present and aware of the cool water around her. (Subtle intake of breath). "Got it. The prize is mine. That bastard’s got nothing now." (A soft chuckle). "Just listen to them patrolling so close…" (A pause, a slight shift in tone). "Almost feel sorry for them." (A hint of dark satisfaction). CUT TO: The same voice now fractured, gasping, raw with panic. (A choked, desperate sound). "No—no no no—can’t—can’t breathe—they see…" (A ragged inhale). "Gotta stay—down—shut up—shut—" (A strangled whisper). "No—" (Hyperventilating inhales). "Can't—breathe—" (Muffled, desperate exhalations). "They see—" (A broken, choked sound). "Shut—" (A panicked, fragmented vocalization). The performance across both moments feels like a plummet from serene confidence into suffocating terror, the contrast echoing the sudden shift from control to utter vulnerability.
🏆
reward 2.37invWER 1.00genu 0.78blend 4.59dur 13sseed 2
ASR heard: “Got it. The prize is mine. That bastard's got nothing now. Just listen to them patrolling so close. Almost feel sorry for them. No, no, no, no, can't ”
#2
reward 1.77invWER 0.96genu 0.20blend 4.12dur 12sseed 0
ASR heard: “Got it. The prize is mine. That bastard's got nothing now. Just listen to them patrolling so close. Almost feel sorry for them. No, no, no, no. Can't ”
#3
reward 1.70invWER 0.60genu 1.77blend 3.44dur 15sseed 1
ASR heard: “Almost feel sorry for them. No, no, no, no, can't can't breathe, they see. Gotta stay, down, shut up, shut. No, can't breathe, they see, shut.”
#4
reward 0.91invWER 0.96genu 0.22blend 0.71dur 16sseed 3
ASR heard: “Got it. The prize is mine. That bastard's got nothing now. Just listen to them patrolling so close. Almost feel sorry for them. No, no, no, no, can't ”
EnglishExtreme (Physical)group g0127 · 4 takes
32 Female with a warm, slightly husky timbre, delivering this high-quality studio voice recording with no background noise. The same voice is quietly observing movement, a subtle tension beneath a calm surface. (Slight inhale) "You always drive like you own the whole damn road." (Gentle sigh) "Honestly, it’s… it’s reckless. But… beautiful." CUT TO: The same voice now fractured, raw, and overwhelmed by sudden, violent impact. (Sharp intake of breath) "Wha… oh god… the noise…" (Gasping, ragged breaths) "I can’t hear… I can't… is that blood on your face?" (A choked, broken sob) "Oh God, oh God…" (A long, wet moan trails off) The performance across both moments should feel like a horrifying, instantaneous shift from intimate observation to primal terror.
🏆
reward 2.75invWER 0.97genu 1.84blend 3.77dur 12sseed 3
ASR heard: “You always drive like you own the whole damn road. Honestly, it's it's reckless, but beautiful. What oh God, the noise I can't hear, I can't is that b”
#2
reward 2.56invWER 0.97genu 1.15blend 4.33dur 15sseed 1
ASR heard: “You always drive like you own the whole damn road. Honestly, it's it's reckless. But beautiful. Well oh God, the noise I can't hear. I can't. Is that ”
#3
reward 0.44invWER 0.03genu 0.28blend 3.09dur 49sseed 2
ASR heard: “Cut to the same voice now fractured, raw, and overwhelmed by sudden violent impact. Gasping, ragged breath, soul line all lent moan trails off.”
#4
reward 0.36invWER 0.03genu 0.84blend 2.13dur 18sseed 0
ASR heard: “Gasping ragged breaths, a long, wet moan trails off. The performance across both moments should feel like a horrifying instantaneous shift from intima”
EnglishExtreme (Physical)group g0128 · 4 takes
A 78-year-old woman with a warm, slightly raspy timbre, delivering this high-quality studio voice recording with no background noise. The same voice is calm, sensual, perceptive — the actor is present and aware of the warmth in her hands and the quiet of the morning. "There now. That's better. The chill doesn't stand a chance against this." "Just right." (Short direction) "Soft, slow, contented words." (Short direction) "A gentle sigh." CUT TO: The same voice now raw and fractured — the extreme physical sensation has struck, vocal cords straining. "You think—" (grunt as you push back) "—I've never—" (breath saws, voice cracks with effort) "—had to hold on before?" "No." "Not this." "Don't you dare." (Short direction) "Distorted, panicked vocalizations." (Short direction) "A desperate, strangled hiss." The performance across both moments should feel like a world of serenity violently shattered by desperate, primal survival.
🏆
reward 1.30invWER 1.00genu 0.72blend 0.23dur 20sseed 3
ASR heard: “There now, that's better. The chill doesn't stand a chance against this. Just right. Soft, slow, contented words. A gentle sigh you think I've never h”
#2
reward 1.12invWER 0.00genu 0.61blend 4.04dur 55sseed 2
#3
reward 1.04invWER 0.04genu 1.26blend 0.09dur 15sseed 1
ASR heard: “The same voice now raw and fractured, the extreme physical sensation has struck, vocal cords straining. Ah but performance across both moments should ”
#4
reward 0.88invWER 0.02genu 0.52blend 3.47dur 55sseed 0
ASR heard: “Pfy in the lie and the lie in my hell liemo on the lie and the lie and mo Hinalaya Mani Dayler Hindu Papa Big Suser Chusha Disappi Ruby and Mr. Afatte”
EnglishExtreme (Physical)group g0129 · 4 takes
A 78-year-old woman with a warm, slightly gravelly timbre, delivering this high-quality studio voice recording with no background noise. The same voice is calm, sensual, perceptive, aware of the warmth spreading from the mug in her hands. (Softly) "This is my reward, you know. Fifty years of ice and metal. And it was worth every frozen night." (A gentle sigh) "All the doubts, the setbacks… they seem so distant now." CUT TO: The same voice now fractured, a jagged, chattering whisper fighting against a rising wave of pain. (Shakily) "No... not again... I should've known." (A gasp) "My hands... they're gone. They're not mine." (A broken, whimpering sound) "Cold... so cold..." (A low, animal moan.) The performance across both moments should feel like a devastating plummet from serene triumph into agonizing, irreversible loss.
🏆
reward 2.34invWER 0.50genu 1.49blend 3.83dur 46sseed 3
ASR heard: “Hat thee here of ice and metal, and it was worth every frozen night. All the doubts, the setbacks, they seem so distant now. No, I'm not sure.”
#2
reward 1.39invWER 0.06genu 1.66blend 2.28dur 14sseed 2
ASR heard: “My mo moan. The woman's across both moments should feel like a devastating plummet from serene triumph into agonizing irreversible loss.”
#3
reward 1.28invWER 0.98genu 0.74blend 1.28dur 15sseed 0
ASR heard: “This is my reward, you know. Fifty years of ice and metal, and it was worth every frozen night. All the doubts, the setbacks, they seem so distant now”
#4
reward 1.06invWER 0.98genu 0.38blend 1.43dur 13sseed 1
ASR heard: “This is my reward, you know. Fifty years of ice and metal, and it was worth every frozen night. All the doubts, the setbacks, they seem so distant now”
GermanExtreme (Physical)group g0140 · 4 takes
A 45-year-old female with a warm, slightly husky timbre, delivering this high-quality studio voice recording with no background noise. The same voice is calm, gentle, and deeply attuned to the quiet beauty of the winter landscape – the actor is present, aware of the cold air and the soft crunch of snow. (Softly calling out) "Hallo, mein Liebling. Komm langsam zurueck, das Eis ist schmal in der Mitte." (A slight pause, listening) "Ich hoeere dich nicht gut." CUT TO: The same voice now fractured and raw, echoing with pain and desperation. (Screaming, gasping) "Halt durch— Ich komme— Bleib oben— Kopf oben— Kopf oben!" (Labored breaths, fragmented cries) "Halt durch— Ich komme— Bleib oben— Kopf oben!" (Distorted vocalizations, choked sobs) "Halt… oben…"
🏆
reward 2.97invWER 0.95genu 1.54blend 6.48dur 18sseed 0
ASR heard: “Hallo, mein Liebling, komm langsam zurück, das Eis ist schmal in der Mitte. Ich hure dich nicht gut, halt durch. Ich komme, bleib oben, Kopf oben, Kop”
#2
reward 1.92invWER 0.92genu 0.60blend 6.41dur 51sseed 1
ASR heard: “Hallo, mein Liebling. Komm langsam zu, Roheg. Das Eis ist schmal in der Mitte. Ich hole dich nicht gut. Halt durch, ich komme, bleib oben, Kopf oben, ”
#3
reward 1.30invWER 0.42genu 1.57blend 4.73dur 19sseed 2
ASR heard: “Hallo, mein Liebling, komm langsam zurück. Das Eis ist schmal in der Mitte. Ich hure dich nicht gut. Halt durch. Ich komme bleib oben Kopf, oben Kopf ”
#4
reward 0.62invWER 0.68genu 0.72blend 3.97dur 16sseed 3
ASR heard: “Hallo mein Liebling, komm langsam zurück, das Eis ist schmal in der Mitte, ich hure dich nicht gut, halt durch, ich komme, bleib oben, Kopf oben, Kopf”
GermanExtreme (Physical)group g0141 · 4 takes
A 35-year-old female with a husky, slightly breathy vocal quality, delivering this high-quality studio voice recording with no background noise. The same voice is calm, sensual, perceptive — the actor is present and aware of the cool water around her. (Slight breath) "Ich habe es. Die Beute ist mein. Dieser Bastard hat nichts mehr. Nur lausche ich ihrem Patrouillieren so nah… fast schade um sie." (A soft chuckle) "Hoert zu, wie sie sich so langsam bewegen." "Es ist fast… beruhigend." CUT TO: The same voice now [the extreme physical sensation has struck — voice transforms]. (short direction in English) "Panic, choking, desperate gasps." (short direction in English) "Fragmented, broken vocalizations." The performance across both moments should feel utterly jarring, a stark contrast between quiet triumph and raw, visceral terror. "Nein—nein nein—kann—kann nicht atmen—sie sehen—muss—muss unten bleiben—heimisch—schtopp—" "Atmen—atmen—" (A strangled sob) "Ich—ich kann nicht—" "Luft—brauche Luft—" (Rapid, shallow gasps) "Hilfe—bitte—" "Hoert auf—" (Disjointed, panicked whispers) "Nicht—nicht—"
🏆
reward 2.98invWER 0.61genu 1.66blend 8.85dur 29sseed 0
ASR heard: “Hört zu, wie sie sich so langsam bewegen. Es ist fast beruhigend. Panic, choking, desperate gasps, fragmented, broken vocalizations. Nein, nein, nein,”
#2
reward 2.74invWER 0.50genu 1.61blend 9.01dur 55sseed 3
ASR heard: “Ich hab es. Die Beute ist mine. Dieser Bastard hat nichts mehr. Nur lausche ich ihrem Patrouillieren so nah. Fast schade um sie. Hört zu, wie sie sich”
#3
reward 1.55invWER 0.01genu 1.55blend 6.28dur 55sseed 2
ASR heard: “Night. Night. Um,”
#4
reward 0.00invWER 0.00genu 0.96blend 0.40dur 55sseed 1
ASR heard: “Stra Fages.”

5 · How to run it most efficiently

Model: MOSS-TTS 8B voice-acting (checkpoint-last), no-reference, 24 kHz, sdpa/bf16. Scorers: Parakeet-TDT-0.6b-v3 (invWER), VoiceCLAP-commercial genuineness & blend heads. Benchmark: 200 prompts × 4 seeds = 796 takes. Generated on A100-80GB (GPUs 0 & 4).