V100-PCIE-32GB · 32 GB
Kokoro-82M
Open a report for complete measurements.
- Task
- Speech synthesis · Text → Audio
- Workload
- Text → Audio · kokoro-english-v1 · Test duration 6 s
- Batch size
- Batch 2
- Format · Engine
- fp32 · kokoro · PyTorch 2.12.1+cu126
- Unit
- audio-sec/sec
- Mean throughput
- 37.8audio-sec/sec
- Observed GPU occupancy
- 1.5 GiB
- P95 latency
- 237 ms
- Reports combined
- 11
Launch this model
This engine does not yet have an executable serving recipe.
Individual tests
Text → Audio · Speech synthesis · fp32 · kokoro · tts · audio-sec/sec
Sortable table
| Test date | Throughput | Concurrency / batch | P95 (s) | TTFT (ms) | Observed occupancy (GiB) | GPU power (W) | Decode (tok/s) | Uncached prefill (tok/s) | Parameters | View report |
|---|---|---|---|---|---|---|---|---|---|---|
| 10/8/2026, 3:49:37 AM | 40.1 audio-sec/sec | 2 | 0.218 | — | 1.45 | 85.25 | — | Uncached prefill unavailableView report | Parameters ↗ | baa57457 ↗ |
| 10/8/2026, 3:58:04 AM | 38 audio-sec/sec | 2 | 0.226 | — | 1.45 | 89.39 | — | Uncached prefill unavailableView report | Parameters ↗ | 5a2ffbba ↗ |
| 10/8/2026, 4:02:22 AM | 38.89 audio-sec/sec | 2 | 0.227 | — | 1.45 | 85.98 | — | Uncached prefill unavailableView report | Parameters ↗ | 06e33423 ↗ |
| 10/8/2026, 4:06:57 AM | 29.89 audio-sec/sec | 2 | 0.286 | — | 1.45 | 70.89 | — | Uncached prefill unavailableView report | Parameters ↗ | c86a1ba5 ↗ |
| 10/8/2026, 4:11:08 AM | 39.92 audio-sec/sec | 2 | 0.217 | — | 1.45 | 82.76 | — | Uncached prefill unavailableView report | Parameters ↗ | cfff852e ↗ |
| 10/8/2026, 4:16:07 AM | 36.73 audio-sec/sec | 2 | 0.235 | — | 1.45 | 78.93 | — | Uncached prefill unavailableView report | Parameters ↗ | 1d979fbe ↗ |
| 10/8/2026, 4:20:29 AM | 37.48 audio-sec/sec | 2 | 0.238 | — | 1.45 | 91.09 | — | Uncached prefill unavailableView report | Parameters ↗ | 31274344 ↗ |
| 10/8/2026, 4:24:56 AM | 38.76 audio-sec/sec | 2 | 0.221 | — | 1.45 | 79.5 | — | Uncached prefill unavailableView report | Parameters ↗ | d21a9695 ↗ |
| 10/8/2026, 4:29:45 AM | 39.01 audio-sec/sec | 2 | 0.227 | — | 1.45 | 86.32 | — | Uncached prefill unavailableView report | Parameters ↗ | 0d92eee3 ↗ |
| 10/8/2026, 4:35:06 AM | 37.94 audio-sec/sec | 2 | 0.262 | — | 1.45 | 76.93 | — | Uncached prefill unavailableView report | Parameters ↗ | 61a6e3a1 ↗ |
| 10/8/2026, 4:39:35 AM | 39.37 audio-sec/sec | 2 | 0.249 | — | 1.45 | 81.27 | — | Uncached prefill unavailableView report | Parameters ↗ | b40f8533 ↗ |
Recorded test parameters
Configuration 1 · Parameters
- model
- hexgrad/Kokoro-82M
- engine
- kokoro
- quant
- fp32
- method
- audio-native-v1
- task
- tts
- dataset
- kokoro-english-v1
- context
- tts
- input_tokens
- 0
- output_tokens
- 0
- duration
- 6
- concurrency
- 2
- unit
- audio-sec/sec
- software
- { "torch": "2.12.1+cu126", "transformers": "4.57.6", "ctranslate2": "4.8.2", "faster-whisper": "1.2.1", "kokoro": "0.9.4", "OMP_NUM_THREADS": "6", "MKL_NUM_THREADS": "6", "OPENBLAS_NUM_THREADS": "1", "NUMEXPR_NUM_THREADS": "4" }
- engine_provenance
- —
- kernel_settings
- —
- precision
- —
- resolution
- —
- performance_task
- —
- performance_scope
- —
- batch_size
- —
- dimensions
- —
- weight_quantization
- —
- max_sequence_length
- —
- model_scope
- —
- performance_dataset
- —
- workload_mode
- —
- graph_mode
- —
- model_revision
- —
- client_version
- python-native-bench-1.1.3
Configuration 2 · Parameters
- model
- hexgrad/Kokoro-82M
- engine
- kokoro
- quant
- fp32
- method
- audio-native-v1
- task
- tts
- dataset
- kokoro-english-v1
- context
- tts
- input_tokens
- 0
- output_tokens
- 0
- duration
- 6
- concurrency
- 2
- unit
- audio-sec/sec
- software
- { "torch": "2.12.1+cu126", "transformers": "4.57.6", "ctranslate2": "4.8.2", "faster-whisper": "1.2.1", "kokoro": "0.9.4", "OMP_NUM_THREADS": "4", "MKL_NUM_THREADS": "4", "OPENBLAS_NUM_THREADS": "1", "NUMEXPR_NUM_THREADS": "4" }
- engine_provenance
- —
- kernel_settings
- —
- precision
- —
- resolution
- —
- performance_task
- —
- performance_scope
- —
- batch_size
- —
- dimensions
- —
- weight_quantization
- —
- max_sequence_length
- —
- model_scope
- —
- performance_dataset
- —
- workload_mode
- —
- graph_mode
- —
- model_revision
- —
- client_version
- python-native-bench-1.1.3