Speech-to-Text
LiveHow each system turns speech to text for a live agent — how fast it ends the turn in a live call, accuracy on clean English, and price. Streaming models a caller actually hears rank first; batch transcription APIs sit at the end. Measured through one gateway, no human scoring.
| Rank · Provider / Model | |||
|---|---|---|---|
01 Smallest AI Pulse | 5.1% | 183ms †Flushed at end-of-speech: Speko forces the final on the turn signal for flush-capable providers, instead of waiting the vendor silence timer (~1–3s).n=22 | — |
02 Smallest AI Pulse Pro | 6.4%95% CI 4.4–8.8 · n=50 | —Batch / pre-recorded HTTP model — no streaming socket, so it cannot carry a live turn (like the other batch APIs). | — |
03 Deepgram Nova-3 | 9.8% | 368msn=19 | $0.0048 |
04 Deepgram Nova-2 | 13.9% | 387msn=18 | $0.0043 |
05 Cartesia Ink-2 | 11.0% | 575msn=22 | $0.0022 |
06 Alibaba Qwen3-ASR | 2.8% | 710msn=12 | — |
07 Google Chirp 3 | 3.9% | 995msn=14 | — |
08 ElevenLabs Scribe v2 | 3.1% | 1750msn=14 | $0.0067 |
09 Soniox stt-rt-v5 | 7.5%95% CI 4.8–10.7 · n=50 | 269msus-east4 p50 over the 11/20 clips that finalized on Soniox NATIVE endpointing (flush off on staging). 9/20 did not finalize from silence alone — native turn-detection is unreliable; the flush path (on in prod) is needed for a dependable final, where Soniox should reach the ~180ms flush-forced class like Smallest. 5/11 were early finalizes.n=11/20 · native | $0.002 |
10 OpenAI GPT-4o Transcribe | 2.3% | — | $0.006 |
11 OpenAI GPT-4o-mini Transcribe | 2.7% | — | $0.003 |
12 ElevenLabs Scribe v1 | 3.1% | — | $0.0067 |
13 OpenAI Whisper-1 | 3.9% | — | $0.006 |
14 xAI Grok STT | 4.8% | — | — |
15 Gradium Gradium ASR | 13.7% | — | — |
16 Cartesia Ink-Whisper | 25.2% | — | $0.0022 |
For a live agent the number a caller actually feels is end-of-turn latency, not clean-read WER — so this board leads with the models that stream a live turn through the gateway, ordered by how fast they finalize it. Smallest Pulse ends the turn in ~183 ms (flushed at end-of-speech); Deepgram nova-3/nova-2 and Cartesia ink-2 sit in a 368–575 ms band; Alibaba Qwen3-ASR lands at ~710 ms — fast, and at 2.8% WER, though its DashScope endpoint is Asia-hosted; Google Chirp 3 trails at ~995 ms; and ElevenLabs Scribe v2 sits near 1.75 s on its patient default VAD (tunable, not a floor). Soniox streams too, but its live end-of-turn is not yet measured. The batch transcription APIs below (OpenAI, ElevenLabs Scribe v1, xAI Grok, Gradium) win clean-read WER — a near-solved 2.3–3.9% cluster — but do not stream a live turn here, so end-of-turn reads as a dash; click WER to rank the whole board by accuracy instead. Caveats: streaming-first ASRs like nova-3 and ink-2 score worse on whole-clip WER because batch scoring truncates them (treat as a lower bound), and Grok and Gradium were measured on loudness-normalized input (see flags).