Deepgram vs Speechmatics
Deepgram Nova-3 vs Speechmatics Enhanced for transcription. WER: 5.2% vs 4% — edge Speechmatics Langs: 50 vs 70 — edge Speechmatics Latency: 300 ms vs 500 ms — edge Deepgram Computed from public benchmarks with dated sources; updated 2026-08-20.
Head to head
| Axis | Deepgram Nova-3 | Speechmatics Enhanced | Edge |
|---|---|---|---|
| Price | $4.3 per 1000 min | —* | — |
| WER | 5.2% | 4% | Speechmatics |
| Langs | 50 | 70 | Speechmatics |
| Latency | 300 ms | 500 ms | Deepgram |
| diarization | ✓ | ✓ | — |
| timestamps | ✓ | ✓ | — |
| vocab | ✓ | ✓ | — |
* token-/credit-priced — the headline understates real per-unit cost, so no price edge is awarded.
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.SpeechmaticsLong-standing broad multilingual coverage (~70 languages) with diarization, timestamps and a custom dictionary. Vendor no longer publishes per-model rates (the pricing page defers to a portal), so the June 2026 $6.70/1000 min could not be re-verified and third-party figures conflict up to 2-3x higher; the price is therefore marked not comparable. Mid-pack accuracy (4.0% WER).
Sources
- Artificial Analysis — Speech to Text2026-08-20
- Open ASR Leaderboard2026-06-19