32 GB · 25 reports

V100-PCIE-32GB · 32 GB

Qualified results first; original reports and full tables stay below.

Speech synthesis · Qualified results

Speeds compare only within one exact test. Recipe availability is checked on each report against the published recipe executor.

Kokoro-82M
hexgrad · 2 tested configurations
- fp32 · Text → Audio · kokoro-english-v1 · Test duration 6 s
  Throughput · Batch 1
  35.8audio-sec/sec
  Best measured · Batch 2
  37.6audio-sec/sec
  Observed GPU occupancy
  1.4GiB
  P95 latency
  105 ms
  Evidence report and run availability ↗Same test on other GPUs
  Test details · 2 batch sizes
  Occupancy includes engine buffers and reserved KV cache. It is not the minimum VRAM required by the model.
  Reports combined counts reports, not inference requests. Request successes and failures are listed in each report.
  Engine
  kokoro · PyTorch 2.12.1+cu126
  Protocol
  audio-native-v1
  Dataset
  kokoro-english-v1
  Workload
  tts
  Batch size Mean throughput (audio-sec/sec) P95 TTFT Observed occupancy (GiB) GPU power (W) Reports combined View report
  B1 35.82 105 ms — 1.45 76.2 10 b40f8533 ↗
  B2 37.6 239 ms — 1.45 82.3 10 b40f8533 ↗
- 1 other test configurations on this GPU
  - fp32 · Text → Audio · kokoro-english-v1 · Test duration 6 s
    Throughput · Batch 1
    37.2audio-sec/sec
    Best measured · Batch 2
    40.1audio-sec/sec
    Observed GPU occupancy
    1.4GiB
    P95 latency
    100 ms
    Evidence report and run availability ↗Same test on other GPUs
    Test details · 2 batch sizes
    Occupancy includes engine buffers and reserved KV cache. It is not the minimum VRAM required by the model.
    Reports combined counts reports, not inference requests. Request successes and failures are listed in each report.
    Engine
    kokoro · PyTorch 2.12.1+cu126
    Protocol
    audio-native-v1
    Dataset
    kokoro-english-v1
    Workload
    tts
    Batch size Mean throughput (audio-sec/sec) P95 TTFT Observed occupancy (GiB) GPU power (W) Reports combined View report
    B1 37.23 100 ms — 1.45 77.6 1 baa57457 ↗
    B2 40.1 218 ms — 1.45 85.3 1 baa57457 ↗

Batch size	Mean throughput (audio-sec/sec)	P95	TTFT	Observed occupancy (GiB)	GPU power (W)	Reports combined	View report
B1	35.82	105 ms	—	1.45	76.2	10	b40f8533 ↗
B2	37.6	239 ms	—	1.45	82.3	10	b40f8533 ↗

Batch size	Mean throughput (audio-sec/sec)	P95	TTFT	Observed occupancy (GiB)	GPU power (W)	Reports combined	View report
B1	37.23	100 ms	—	1.45	77.6	1	baa57457 ↗
B2	40.1	218 ms	—	1.45	85.3	1	baa57457 ↗

Occupancy includes engine buffers and reserved KV cache. It is not the minimum VRAM required by the model.

Sortable tables by exact test condition · 4 rows

C: concurrency · B: batch · —: not measured

Speech synthesis · Text → Audio · kokoro-english-v1 · Test duration 6 s · Batch 1 · audio-sec/sec · 1 rows
Model	Format	Engine	Workload	Concurrency / batch	Mean throughput	Best measured	TTFT (ms)	P95 (s)	Observed occupancy (GiB)	GPU power (W)	Lowest measured	Throughput / W	Reports combined
hexgrad/Kokoro-82MScaling curves ↗	fp32	kokoro	ttsaudio-native-v1	B1	35.82	37.2	—	0.105	1.45	76.2	31.11	0.47	10 ↗

Speech synthesis · Text → Audio · kokoro-english-v1 · Test duration 6 s · Batch 1 · audio-sec/sec · 1 rows
Model	Format	Engine	Workload	Concurrency / batch	Mean throughput	Best measured	TTFT (ms)	P95 (s)	Observed occupancy (GiB)	GPU power (W)	Lowest measured	Throughput / W	Reports combined
hexgrad/Kokoro-82MScaling curves ↗	fp32	kokoro	ttsaudio-native-v1	B1	37.23	37.23	—	0.1	1.45	77.6	37.23	0.48	1 ↗

Speech synthesis · Text → Audio · kokoro-english-v1 · Test duration 6 s · Batch 2 · audio-sec/sec · 1 rows
Model	Format	Engine	Workload	Concurrency / batch	Mean throughput	Best measured	TTFT (ms)	P95 (s)	Observed occupancy (GiB)	GPU power (W)	Lowest measured	Throughput / W	Reports combined
hexgrad/Kokoro-82MScaling curves ↗	fp32	kokoro	ttsaudio-native-v1	B2	37.6	39.92	—	0.239	1.45	82.3	29.89	0.46	10 ↗

Speech synthesis · Text → Audio · kokoro-english-v1 · Test duration 6 s · Batch 2 · audio-sec/sec · 1 rows
Model	Format	Engine	Workload	Concurrency / batch	Mean throughput	Best measured	TTFT (ms)	P95 (s)	Observed occupancy (GiB)	GPU power (W)	Lowest measured	Throughput / W	Reports combined
hexgrad/Kokoro-82MScaling curves ↗	fp32	kokoro	ttsaudio-native-v1	B2	40.1	40.1	—	0.218	1.45	85.3	40.1	0.47	1 ↗

Scaling curves and individual runs

One model and test protocol per curve; different workloads are never joined into one line.

hexgrad/Kokoro-82M · fp32 · kokoro · tts · audio-native-v1

Mean throughput audio-sec/sec

Measured points only. Select a point to open its source report.

Mean run P95 latency s

Measured points only. Select a point to open its source report.

Mean GPU memory used GB

Measured points only. Select a point to open its source report.

Each point is the mean of recorded runs at that level. A source report may differ from the mean; all runs are listed below.

Runs behind the curve · hexgrad/Kokoro-82M · audio-sec/sec · 20 rows
Batch size	Throughput	Decode (tok/s)	Uncached prefill (tok/s)	GPU power (W)	TTFT (ms)	P95 (s)	Observed occupancy (GiB)	Test date	View report
1	36.54 audio-sec/sec	—	Uncached prefill unavailableView report	73.17	—	0.103	1.45	10/8/2026, 4:39:35 AM	b40f8533 ↗
1	36.4 audio-sec/sec	—	Uncached prefill unavailableView report	77.89	—	0.097	1.45	10/8/2026, 4:35:06 AM	61a6e3a1 ↗
1	36.47 audio-sec/sec	—	Uncached prefill unavailableView report	82.68	—	0.101	1.45	10/8/2026, 4:29:45 AM	0d92eee3 ↗
1	36.15 audio-sec/sec	—	Uncached prefill unavailableView report	72.55	—	0.108	1.45	10/8/2026, 4:24:56 AM	d21a9695 ↗
1	36.46 audio-sec/sec	—	Uncached prefill unavailableView report	71.22	—	0.095	1.45	10/8/2026, 4:20:29 AM	31274344 ↗
1	35 audio-sec/sec	—	Uncached prefill unavailableView report	77.42	—	0.11	1.45	10/8/2026, 4:16:07 AM	1d979fbe ↗
1	37.2 audio-sec/sec	—	Uncached prefill unavailableView report	78.61	—	0.096	1.45	10/8/2026, 4:11:08 AM	cfff852e ↗
1	31.11 audio-sec/sec	—	Uncached prefill unavailableView report	75.39	—	0.148	1.45	10/8/2026, 4:06:57 AM	c86a1ba5 ↗
1	36.42 audio-sec/sec	—	Uncached prefill unavailableView report	77.63	—	0.098	1.45	10/8/2026, 4:02:22 AM	06e33423 ↗
1	36.48 audio-sec/sec	—	Uncached prefill unavailableView report	75.46	—	0.097	1.45	10/8/2026, 3:58:04 AM	5a2ffbba ↗
2	37.94 audio-sec/sec	—	Uncached prefill unavailableView report	76.93	—	0.262	1.45	10/8/2026, 4:35:06 AM	61a6e3a1 ↗
2	37.48 audio-sec/sec	—	Uncached prefill unavailableView report	91.09	—	0.238	1.45	10/8/2026, 4:20:29 AM	31274344 ↗
2	39.92 audio-sec/sec	—	Uncached prefill unavailableView report	82.76	—	0.217	1.45	10/8/2026, 4:11:08 AM	cfff852e ↗
2	38.89 audio-sec/sec	—	Uncached prefill unavailableView report	85.98	—	0.227	1.45	10/8/2026, 4:02:22 AM	06e33423 ↗
2	38 audio-sec/sec	—	Uncached prefill unavailableView report	89.39	—	0.226	1.45	10/8/2026, 3:58:04 AM	5a2ffbba ↗
2	39.37 audio-sec/sec	—	Uncached prefill unavailableView report	81.27	—	0.249	1.45	10/8/2026, 4:39:35 AM	b40f8533 ↗
2	39.01 audio-sec/sec	—	Uncached prefill unavailableView report	86.32	—	0.227	1.45	10/8/2026, 4:29:45 AM	0d92eee3 ↗
2	38.76 audio-sec/sec	—	Uncached prefill unavailableView report	79.5	—	0.221	1.45	10/8/2026, 4:24:56 AM	d21a9695 ↗
2	36.73 audio-sec/sec	—	Uncached prefill unavailableView report	78.93	—	0.235	1.45	10/8/2026, 4:16:07 AM	1d979fbe ↗
2	29.89 audio-sec/sec	—	Uncached prefill unavailableView report	70.89	—	0.286	1.45	10/8/2026, 4:06:57 AM	c86a1ba5 ↗

Original reports

Every report for this task, including quality, diagnostic and legacy runs that do not enter comparisons.

Original reports · 11 rows
Model	Test task	Format	Engine	Workload	Protocol	Quality score	Test date
Kokoro-82M ↗	Speech synthesis	fp32	kokoro	tts	audio-native-v1	—	10/8/2026, 4:39:35 AM
Kokoro-82M ↗	Speech synthesis	fp32	kokoro	tts	audio-native-v1	—	10/8/2026, 4:35:06 AM
Kokoro-82M ↗	Speech synthesis	fp32	kokoro	tts	audio-native-v1	—	10/8/2026, 4:29:45 AM
Kokoro-82M ↗	Speech synthesis	fp32	kokoro	tts	audio-native-v1	—	10/8/2026, 4:24:56 AM
Kokoro-82M ↗	Speech synthesis	fp32	kokoro	tts	audio-native-v1	—	10/8/2026, 4:20:29 AM
Kokoro-82M ↗	Speech synthesis	fp32	kokoro	tts	audio-native-v1	—	10/8/2026, 4:16:07 AM
Kokoro-82M ↗	Speech synthesis	fp32	kokoro	tts	audio-native-v1	—	10/8/2026, 4:11:08 AM
Kokoro-82M ↗	Speech synthesis	fp32	kokoro	tts	audio-native-v1	—	10/8/2026, 4:06:57 AM
Kokoro-82M ↗	Speech synthesis	fp32	kokoro	tts	audio-native-v1	—	10/8/2026, 4:02:22 AM
Kokoro-82M ↗	Speech synthesis	fp32	kokoro	tts	audio-native-v1	—	10/8/2026, 3:58:04 AM
Kokoro-82M ↗	Speech synthesis	fp32	kokoro	tts	audio-native-v1	—	10/8/2026, 3:49:37 AM

Running benchmarks

0Updates automatically

No benchmarks are currently reported as running.

Methodology

Average of error-free runs for the selected recorded configuration. Missing methods and fallback runs are not ranked. Performance index = average / fastest matching hardware × 100. Value index = performance per purchase cost / best comparable value × 100. Quality is separate by dataset and protocol. Prices are dated seller listings, not completed transactions or market averages.

Updated 11/5/2026, 11:07:29 PM