Skip to content
m.recipes
Find a configuration
Sign in
12 GB · 4 reports

RTX 3080 Ti · 12 GB

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

All tasks · Qualified results

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

  1. Kokoro-82M

    hexgrad · 1 tested configuration

    • fp32 · Text → Audio · kokoro-english-v1 · Test duration 6 s
      Throughput · Batch 1
      49.2audio-sec/sec
      Best measured · Batch 2
      52.8audio-sec/sec
      Observed GPU occupancy
      1GiB
      P95 latency
      73 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+cu130
      Protocol
      audio-native-v1
      Dataset
      kokoro-english-v1
      Workload
      tts
      Batch sizeMean throughput (audio-sec/sec)P95TTFTObserved occupancy (GiB)GPU power (W)Reports combinedView report
      B149.1873 ms—1158.913c7db092 ↗
      B252.8140 ms—1.06194.713c7db092 ↗

  2. SenseVoiceSmall

    iic · 1 tested configuration

    • fp16 · Audio → Text · asr_prompts
      Throughput · Batch 1
      360audio-sec/sec
      Best measured · Batch 2
      365audio-sec/sec
      Observed GPU occupancy
      1.3GiB
      P95 latency
      Not recorded
      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
      funasr · PyTorch 2.12.1+cu130
      Protocol
      audio-native-v1
      Dataset
      asr_prompts
      Workload
      asr
      Batch sizeMean throughput (audio-sec/sec)P95TTFTObserved occupancy (GiB)GPU power (W)Reports combinedView report
      B1359.69——1.27101.71c6b5f31d ↗
      B2364.87——1.27101.61c6b5f31d ↗

  3. faster-whisper-tiny

    Systran · 1 tested configuration

    • fp16 · Audio → Text · asr_prompts · Test duration 2 s
      Throughput · Batch 1
      65.7audio-sec/sec
      Best measured · Batch 4
      91.8audio-sec/sec
      Observed GPU occupancy
      0.4GiB
      P95 latency
      Not recorded
      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
      faster-whisper · PyTorch 2.12.1+cu130
      Protocol
      audio-native-v1
      Dataset
      asr_prompts
      Workload
      asr
      Batch sizeMean throughput (audio-sec/sec)P95TTFTObserved occupancy (GiB)GPU power (W)Reports combinedView report
      B165.65——0.41134.913f8b97e6 ↗
      B491.84——0.4143.613f8b97e6 ↗

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

Sortable tables by exact test condition · 6 rows

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

Speech recognition · Audio → Text · asr_prompts · Batch 1 · audio-sec/sec · 1 rows
ModelFormatEngineWorkloadConcurrency / batchMean throughputBest measuredTTFT (ms)P95 (s)Observed occupancy (GiB)GPU power (W)Lowest measuredThroughput / WReports combined
iic/SenseVoiceSmallScaling curves ↗fp16funasrasraudio-native-v1B1359.69359.69——1.27101.7359.693.541 ↗
Speech recognition · Audio → Text · asr_prompts · Batch 2 · audio-sec/sec · 1 rows
ModelFormatEngineWorkloadConcurrency / batchMean throughputBest measuredTTFT (ms)P95 (s)Observed occupancy (GiB)GPU power (W)Lowest measuredThroughput / WReports combined
iic/SenseVoiceSmallScaling curves ↗fp16funasrasraudio-native-v1B2364.87364.87——1.27101.6364.873.591 ↗
Speech recognition · Audio → Text · asr_prompts · Test duration 2 s · Batch 1 · audio-sec/sec · 1 rows
ModelFormatEngineWorkloadConcurrency / batchMean throughputBest measuredTTFT (ms)P95 (s)Observed occupancy (GiB)GPU power (W)Lowest measuredThroughput / WReports combined
Systran/faster-whisper-tinyScaling curves ↗fp16faster-whisperasraudio-native-v1B165.6565.65——0.41134.965.650.491 ↗
Speech recognition · Audio → Text · asr_prompts · Test duration 2 s · Batch 4 · audio-sec/sec · 1 rows
ModelFormatEngineWorkloadConcurrency / batchMean throughputBest measuredTTFT (ms)P95 (s)Observed occupancy (GiB)GPU power (W)Lowest measuredThroughput / WReports combined
Systran/faster-whisper-tinyScaling curves ↗fp16faster-whisperasraudio-native-v1B491.8491.84——0.4143.691.840.641 ↗
Speech synthesis · Text → Audio · kokoro-english-v1 · Test duration 6 s · Batch 1 · audio-sec/sec · 1 rows
ModelFormatEngineWorkloadConcurrency / batchMean throughputBest measuredTTFT (ms)P95 (s)Observed occupancy (GiB)GPU power (W)Lowest measuredThroughput / WReports combined
hexgrad/Kokoro-82MScaling curves ↗fp32kokorottsaudio-native-v1B149.1849.18—0.0731158.949.180.311 ↗
Speech synthesis · Text → Audio · kokoro-english-v1 · Test duration 6 s · Batch 2 · audio-sec/sec · 1 rows
ModelFormatEngineWorkloadConcurrency / batchMean throughputBest measuredTTFT (ms)P95 (s)Observed occupancy (GiB)GPU power (W)Lowest measuredThroughput / WReports combined
hexgrad/Kokoro-82MScaling curves ↗fp32kokorottsaudio-native-v1B252.852.8—0.141.06194.752.80.271 ↗
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
Batch size 1: 49.2 audio-sec/sec; Batch size 2: 52.8 audio-sec/sec014.52943.658.112Batch size · log₂

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

Mean run P95 latency s
Batch size 1: 0.073 s; Batch size 2: 0.14 s00.03840.07680.1150.15412Batch size · log₂

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

Mean GPU memory used GB
Batch size 1: 1 GB; Batch size 2: 1.06 GB00.2930.5860.8781.1712Batch size · log₂

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 · 2 rows
Batch sizeThroughputDecode (tok/s)Uncached prefill (tok/s)GPU power (W)TTFT (ms)P95 (s)Observed occupancy (GiB)Test dateView report
149.18 audio-sec/sec—Uncached prefill unavailableView report158.92—0.073110/8/2026, 12:50:51 AM3c7db092 ↗
252.8 audio-sec/sec—Uncached prefill unavailableView report194.71—0.141.0610/8/2026, 12:50:51 AM3c7db092 ↗

Original reports

Every report for this task, including quality, diagnostic and legacy runs that do not enter comparisons.
Original reports · 4 rows
ModelTest taskFormatEngineWorkloadProtocolQuality scoreTest date
SenseVoiceSmall ↗Speech recognitionfp16funasrasraudio-native-v1—10/8/2026, 12:41:46 AM
faster-whisper-tiny ↗Speech recognitionfp16faster-whisperasraudio-native-v1—10/8/2026, 12:41:29 AM
Kokoro-82M ↗Speech synthesisfp32kokorottsaudio-native-v1—10/8/2026, 12:50:51 AM
qwen3.8-flash-next-q2_0 ↗Text generationQ2_0stratallm-128-256strata-native-v1—10/8/2026, 12:57:21 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