PERIT-STT · Leaderboard

Speech recognition, measured against human experts

22 speech-to-text models run on the same real-world audio and scored word by word against transcripts written and reviewed by people.

Back to the overviewPut your model on the board
RankModelMedian LatencyP95 LatencyReal-Time Factor (median)
01
0.7 s
1.3 s
0.06×
02
0.6 s
1.3 s
0.06×
03
1.3 s
1.7 s
0.12×
04
0.7 s
1.5 s
0.06×
05
1.8 s
12.2 s
0.15×
06
1.0 s
1.8 s
0.09×
07
0.8 s
1.1 s
0.07×
08
1.2 s
3.4 s
0.10×
09
3.2 s
16.8 s
0.27×
10
1.0 s
1.8 s
0.09×
11
3.7 s
6.9 s
0.30×
12
5.3 s
11.9 s
0.46×
13
0.4 s
1.3 s
0.04×
14
0.7 s
2.1 s
0.06×
15
2.7 s
5.2 s
0.23×
16
1.0 s
2.0 s
0.09×
17
1.4 s
3.8 s
0.12×
18
1.8 s
3.3 s
0.14×
19
3.6 s
16.7 s
0.29×
20gpt-4o-transcribeOpenAI
1.2 s
2.0 s
0.10×
Accuracy
Word Error Rate (WER)
3.17%#20/22
WER 95% CI (low)
2.86%#21/22
WER 95% CI (high)
3.53%#20/22
Word Accuracy
96.83%#20/22
Character Error Rate (CER)
2.28%#21/22
Strict WER
6.21%#20/22
Match Error Rate (MER)
3.16%#20/22
Word Information Lost (WIL)
3.85%#20/22
Word Information Preserved (WIP)
96.15%#20/22
Error breakdown
Substitution Rate
0.69%#8/22
Deletion Rate
2.12%#22/22
Insertion Rate
0.36%#1/22
Sentence Error Rate (SER)
51.8%#21/22
Exact Match Rate
48.2%#21/22
Length Ratio
0.982×#22/22
Robustness
Mean WER
8.59%#21/22
Median WER
1.28%#21/22
P90 WER
10.40%#22/22
Severe Error Rate
1.0%#19/22
Empty Output Rate
0.0%#1/22
Success Rate
100.0%#1/22
By clip length
WER - short audio (under 5 s)
5.17%#19/22
WER - medium audio (5 to 15 s)
4.66%#22/22
WER - long audio (over 15 s)
2.20%#21/22
Speed
Median Latency
1.2 s#10/22
P95 Latency
2.0 s#10/22
Real-Time Factor (median)
0.10×#10/22
21
3.8 s
7.1 s
0.32×
22
1.3 s
2.8 s
0.11×
Blue marks the best value in each column; a dash means the metric does not apply to that model. Open a model for all 27 metrics and where it places on each.Showing 22 of 22
Best on each axis

The winners, one question at a time.

Most accurate
1.53%Word Error Rate (WER)qwen3-asr-flash-2026-02-10Qwen
Lowest character error
0.87%Character Error Rate (CER)mai-transcribe-1.5Microsoft
Most exact matches
66.1%Exact Match Rateqwen3-asr-flash-2026-02-10Qwen
Most robust
5.72%P90 WERqwen3-asr-flash-2026-02-10Qwen
Fewest dropped words
0.26%Deletion Ratemai-transcribe-1.5Microsoft
Fewest invented words
0.36%Insertion Rategpt-4o-transcribeOpenAI
Fastest
0.4 sMedian Latencynova-3Deepgram
Glossary

What every column means.

Both the reference and the model output go through the same normalization before a single word is counted, so a model is not penalised for writing “2” where the transcriber wrote “two”.

Accuracy

Word- and character-level agreement with the reference, after normalization — and without it.

Word Error Rate (WER)% · lower is better
Headline metric. (substituted + deleted + inserted words) / words in the ground truth, after normalization. Pooled across all clips.
WER 95% CI (low)% · lower is better
Lower bound of the 95% bootstrap confidence interval for WER.
WER 95% CI (high)% · lower is better
Upper bound of the 95% bootstrap confidence interval for WER.
Word Accuracy% · higher is better
100 − WER (floored at 0).
Character Error Rate (CER)% · lower is better
Same as WER but counted on characters; more forgiving of near-miss spellings.
Strict WER% · lower is better
WER with minimal normalization (only lowercase and punctuation removal). Shows how closely raw output matches the ground-truth style.
Match Error Rate (MER)% · lower is better
Errors / (errors + correct words). Bounded at 100% even with heavy hallucination.
Word Information Lost (WIL)% · lower is better
Share of word-level information lost between ground truth and output.
Word Information Preserved (WIP)% · higher is better
100 − WIL.
Error breakdown

What kind of mistake a model makes: the wrong word, a dropped word, or an invented one.

Substitution Rate% · lower is better
Wrong words, as a share of ground-truth words.
Deletion Rate% · lower is better
Missed words (omissions), as a share of ground-truth words.
Insertion Rate% · lower is better
Extra words not spoken (fabrications / hallucinations), as a share of ground-truth words.
Sentence Error Rate (SER)% · lower is better
Share of clips with at least one word error.
Exact Match Rate% · higher is better
Share of clips transcribed perfectly after normalization.
Length Ratiox · closer to 1 is better
Output word count / ground-truth word count. Above 1 suggests hallucination, below 1 suggests skipped speech.
Robustness

How the error is spread across clips — the typical one, the hardest tenth, the outright failures.

Mean WER% · lower is better
Average of per-clip WER (every clip weighted equally).
Median WER% · lower is better
WER of the typical clip.
P90 WER% · lower is better
WER on the hardest 10% of clips; measures robustness.
Severe Error Rate% · lower is better
Share of clips with WER of 50% or more.
Empty Output Rate% · lower is better
Share of clips where the model returned no words.
Success Rate% · higher is better
Share of requests that returned a transcript.
By clip length

WER split by clip duration. Short clips give a model the least context to recover from.

WER - short audio (under 5 s)% · lower is better
WER on short clips.
WER - medium audio (5 to 15 s)% · lower is better
WER on medium-length clips.
WER - long audio (over 15 s)% · lower is better
WER on long clips.
Speed

Request time and real-time factor. Measured under concurrent load, so indicative only.

Median Latencys · lower is better
Median request time. Measured under concurrent load; indicative only.
P95 Latencys · lower is better
95th percentile request time. Measured under concurrent load; indicative only.
Real-Time Factor (median)x · lower is better
Processing time / audio duration. 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 22.

Enter the next runRead the method