Benchmark
Vals VoiceCodeBench
Scores are percentages.
Benchmark sourceScores from Vals AI.
- Models measured
- 10
- Scores recorded
- 10
- Sources
- 1
- IQ weight
- 1.23
Score distribution
- 53%
- Mean
- 54.3%
- Median
- 10.3%
- Std dev
- 44.7%
- 10th percentile
- 65.3%
- 90th percentile
- 30.3%
- Min
This benchmark has a weight of 1.23 for speech models, measured over 10 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.
Leaderboard
| # | Model | Score |
|---|---|---|
| 1 | GPT Live Transcribe· Vals AI | 67.7% |
| 2 | grok-voice-transcribe-2.0· Vals AI | 65.3% |
| 3 | GPT-4o Transcribe· Vals AI | 59.7% |
| 4 | Whisper Large V3 Turbo· Vals AI | 57% |
| 5 | Whisper Large V3· Vals AI | 54.3% |
| 6 | Voxtral Mini Transcribe2602· Vals AI | 53% |
| 7 | grok-voice-transcribe-1.0· Vals AI | 52% |
| 8 | Voxtral Mini Transcribe Realtime· Vals AI | 46% |
| 9 | Cohere Transcribe 03 2026· Vals AI | 44.7% |
| 10 | universal-3-5-pro· Vals AI | 30.3% |