PERIT-ALIGN · Leaderboard
Word alignment, measured against human experts
14 word-alignment models run on the same real-world audio and scored word by word against timestamps marked and reviewed by people.
| Rank | Model | Timing Accuracy @50 ms | Timing Accuracy @100 ms | Boundaries within 20 ms | Boundaries within 50 ms | Boundaries within 100 ms | Boundaries within 200 ms |
|---|---|---|---|---|---|---|---|
| 14 | nemotron-3.5-asr-streaming-multilingual-0.6bNVIDIA | 0.65% | 5.77% | 5.27% | 6.73% | 20.03% | 41.00% |
| 13 | parakeet-tdt-0.6b-v3NVIDIA | 4.16% | 30.96% | 18.69% | 24.73% | 57.19% | 83.46% |
| 12 | whisper-1OpenAI | 15.73% | 45.86% | 18.32% | 42.24% | 69.97% | 89.80% |
| 11 | whisper-large-v3-turboOpenAI | 19.63% | 46.95% | 19.19% | 43.71% | 68.95% | 88.80% |
| 10 | mai-transcribe-2Microsoft | 25.60% | 59.85% | 19.76% | 50.81% | 77.62% | 96.98% |
| 09 | whisper-large-v3OpenAI | 25.79% | 59.89% | 21.70% | 50.59% | 77.61% | 92.05% |
| 08 | nova-3Deepgram | 13.09% | 60.19% | 26.54% | 36.35% | 78.45% | 95.74% |
| 07 | universal-3-5-proAssemblyAI | 22.77% | 63.45% | 20.03% | 48.70% | 80.23% | 96.29% |
| 06 | qwen3-asr-1.7bQwen | 58.75% | 76.60% | 64.79% | 75.52% | 87.56% | 90.15% |
By tolerance
Boundary error
Drift & shape
Coverage & speed
| |||||||
| 05 | qwen3-asr-0.6bQwen | 59.52% | 77.41% | 65.40% | 76.17% | 88.11% | 90.56% |
| 04 | grok-stt-1.0xAI | 38.89% | 78.04% | 28.02% | 62.33% | 88.92% | 97.81% |
| 03 | voxtral-mini-transcribeMistral AI | 43.96% | 85.04% | 28.55% | 67.10% | 93.50% | 98.59% |
| 02 | transcribe-1Fish Audio | 67.69% | 92.33% | 71.13% | 82.82% | 96.93% | 98.71% |
| 01 | Perit Golden Data Fine-tuned ModelPerit | 83.90% | 97.85% | 56.59% | 91.41% | 98.86% | 99.73% |
Blue marks the best value in each column; a dash means the metric does not apply to that model. Open a model for all 19 metrics and where it places on each. Perit's model is a forced aligner: it places the words of a known transcript in time, while the other models transcribe the audio and time the words in one pass.Showing 14 of 14
Best on each axis
The winners, one question at a time.
- Most accurate timing
- 97.85%Timing Accuracy @100 msPerit Golden Data Fine-tuned ModelPerit
- Best speech-to-text timestamps
- 92.33%Timing Accuracy @100 mstranscribe-1Fish Audio
- Tightest timing
- 83.90%Timing Accuracy @50 msPerit Golden Data Fine-tuned ModelPerit
- Smallest boundary error
- 23.8 msBoundary MAEPerit Golden Data Fine-tuned ModelPerit
- Best span overlap
- 84.98%Mean IoUtranscribe-1Fish Audio
- Widest coverage
- 97.70%Word coveragemai-transcribe-2Microsoft
Glossary
What every column means.
Each model's words are matched to the reference words and compared boundary by boundary — start against start, end against end. Coverage and the transcript's own WER are reported next to the timing, so a model cannot look precise by timing only the easy words.
By tolerance
How many words and boundaries land inside each window, from 20 ms to 200 ms.
- Timing Accuracy @50 ms% · higher is better
- The same, at a 50 ms tolerance — how many words a model places tightly, not just roughly.
- Timing Accuracy @100 ms% · higher is better
- Headline metric. Share of reference words whose start and end both land within 100 ms of the reference timing.
- Boundaries within 20 ms% · higher is better
- Share of individual word boundaries (starts and ends counted separately) within 20 ms of the reference.
- Boundaries within 50 ms% · higher is better
- Share of word boundaries within 50 ms of the reference.
- Boundaries within 100 ms% · higher is better
- Share of word boundaries within 100 ms of the reference.
- Boundaries within 200 ms% · higher is better
- Share of word boundaries within 200 ms of the reference.
Boundary error
How far off the boundaries are, in milliseconds — the typical one, the worst tenth, starts against ends.
- Boundary MAEms · lower is better
- Mean absolute gap between predicted and reference word boundaries, starts and ends pooled. The tie-breaker in the ranking.
- Median boundary errorms · lower is better
- Boundary error of the typical boundary.
- P90 boundary errorms · lower is better
- Boundary error at the 90th percentile — how far off the worst tenth of boundaries land.
- Start MAEms · lower is better
- Mean absolute error of word start times alone.
- End MAEms · lower is better
- Mean absolute error of word end times alone.
Drift & shape
Whether a model runs systematically early or late, and whether its word spans have real length.
- Start biasms · closer to 0 is better
- Average signed offset of start times against the reference. Near zero means no systematic drift; the sign shows which way a model leans.
- End biasms · closer to 0 is better
- Average signed offset of end times against the reference.
- Mean IoU% · higher is better
- Average overlap between each word's predicted and reference time span (intersection over union).
- Zero-duration words% · lower is better
- Share of returned words with no length — start and end on the same timestamp.
Coverage & speed
How many words got a timestamp at all, how good the transcript underneath was, and request time — measured under concurrent load, so indicative only.
- Word coverage% · higher is better
- Share of reference words the model returned a timestamp for. Timing is only scored on words that are covered.
- Transcript WER% · lower is better
- Word error rate of the transcript the timestamps came with. Context only — it does not affect the ranking, and a forced aligner, which is given the transcript, has none.
- Success Rate% · higher is better
- Share of requests that returned timestamps.
- Median Latencys · lower is better
- Median request time. Measured under concurrent load; indicative only.
Want your model on this board?
We run it on the same audio, under the same scoring, and publish it next to the other 14.