OpenAIDeepgramAudio matchupSpeech to text
GPT-4o Mini Transcribevs.Nova-3
Which model is better?
GPT-4o Mini Transcribe is slightly ahead in independent tests, but the lead is not certain.
Independent testsArtificial Analysis
GPT-4o Mini Transcribe0 of 1
Nova-30 of 1
Popularityout of 100
GPT-4o Mini Transcribe33
Nova-3ahead51
Weights and access
GPT-4o Mini Transcribeproprietary
Nova-3proprietary
Priceverbatim from the provider
GPT-4o Mini Transcribe$1.25 per 1M audio tokens
Nova-3n/a
Pick GPT-4o Mini Transcribe for
leaning Artificial Analysis
Pick Nova-3 for
more popular
Shared tests
Independent evaluators and leaderboards only, the newest shared snapshot per evaluator. With 95 % ranges a lead counts when they do not overlap; for hit rates, when it holds in a 95 % test. Without any uncertainty figure at most “slightly ahead”.
Artificial Analysis
GPT-4o Mini Transcribe slightly aheadArtificial Analysis AA-WER (non-streaming)
GPT-4o Mini Transcribe4.5 %
Nova-35.2 %
lower is better, as of October 15, 2026
Source · latest October 15, 2026
All data side by side
Prices are verbatim from the provider, because audio models bill per minute, character or token.
| GPT-4o Mini Transcribe | Nova-3 | |
|---|---|---|
| Task | Speech to text | Speech to text |
| Developer | OpenAI | Deepgram |
| Status | Deprecated | Active |
| Weights | proprietary | proprietary |
| Runs locally | no | no |
| Price | $1.25 per 1M audio tokens | n/a |
| Popularity (out of 100) | 33 | 51 |