Deepgram vs OpenAI
Deepgram Nova-3 vs OpenAI gpt-transcribe for transcription. Price: $4.3 per 1000 min vs $4.5 per 1000 min — edge Deepgram WER: 5.2% vs 3.3% — edge OpenAI Computed from public benchmarks with dated sources; updated 2026-08-20.
Head to head
| Axis | Deepgram Nova-3 | OpenAI gpt-transcribe | Edge |
|---|---|---|---|
| Price | $4.3 per 1000 min | $4.5 per 1000 min | Deepgram |
| WER | 5.2% | 3.3% | OpenAI |
| Langs | 50 | — | — |
| Latency | 300 ms | — | — |
| diarization | ✓ | — | — |
| timestamps | ✓ | — | — |
| vocab | ✓ | ✓ | — |
Strengths & caveats
DeepgramStreaming-focused with per-second billing and full diarization, timestamps and keyterm support; markets sub-300 ms latency. Lowest accuracy of the group on the leaderboard (5.2% WER); the latency figure is self-reported, not an independent benchmark. Deepgram now positions a separate, dearer Flux model ($0.0077/min list) as its ultra-low-latency option for voice agents.OpenAIOpenAI's current recommended transcription model: both cheaper ($4.50/1000 min) and more accurate (3.3% WER) than the gpt-4o-transcribe tier it supersedes, with a keywords parameter for biasing terms. Returns plain JSON: no word-level timestamps (only the legacy whisper-1 emits them) and no native diarization — OpenAI ships a separate gpt-4o-transcribe-diarize model for speaker labels. Also billed per token as an alternative to the per-minute rate.
Sources
- Artificial Analysis — Speech to Text2026-08-20
- Open ASR Leaderboard2026-06-19