whatmodel.sh / stt / Soniox v5 Real-Time
Soniox v5 Real-Time
↗ share
+ contribute
Soniox v5 Real-Time is a closed-weights speech-to-text model from Soniox, released June 2026. Licence: proprietary.
Catalogue identity notes
Soniox real-time model stt-rt-v5, released June 16, 2026, replacing stt-rt-v4 (Soniox models changelog). Distinct from the catalogue's async model soniox-v5-async (stt-async-v5, June 11, 2026).
Score within the Open ASR English short-form WER (March 2026) comparison set only.
$ how it compares
The speech-to-text models ranked either side of Soniox v5 Real-Time, out of 176 with a score, then more from Soniox.
| # | model | score | where | ||
|---|---|---|---|---|---|
| 7 |  Ink-2 | 68±3 | api $0.0065/audio min | ||
| 8 |  Universal-3.5 Pro | 67±3 | api $0.0035/audio min | ||
| 9 | Soniox v5 Real-Time | 67±3 | api $0.0020/audio min | ||
| 10 | Soniox v4 Real-Time | 67±3 | comparison only | ||
| 11 |  Gemini 3.5 Transcribe | 67±6 | api $0.0050/audio min | ||
| more from Soniox | |||||
| 6 | Soniox v5 Async | 68±5 | api $0.0017/audio min | ||
see all speech-to-text models · find the ones that fit your machine
$ what it's good at
other tests1 not in a task · 96% of its score
| test | result | vs the others in its run | counts |
|---|---|---|---|
| Pipecat STT Benchmark · pooled semantic WER (1000 smart-turn samples, Sonnet 5.5 judge) | 1.1% ✎ | ▃·▃▅·▅▅·▃▅██ beats 10/16 | 96% |
$ where to run it
Through an API
api · $0.0020/audio min · soniox.com ↗ · price sourceUSD per input audio minute: Soniox bills tokens but itself publishes Speech-to-Text at '$0.12/hour for real-time (streaming)' (pricing page; the comparison widget lists 'stt-rt-v5 ... $0.12/hr'), ÷60. Product: Soniox Real-Time STT API. API model: stt-rt-v5. Settings: real-time streaming. Price basis: vendor-published hourly equivalent, per-hour to per-minute scaling only. Exclusions: async ($0.10/hr), translation/custom token usage. Currency: USD. Soniox official pricing page, observed 2026-10-09.
$ how the score was worked out · Pipecat STT Benchmark · seman… 96%
Each result implies a score on its own, after adjusting for how hard its test is and who ran it. The score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its counts column above; here it's added up by who ran the tests. How it works
| source | results | share of the score |
|---|---|---|
 Pipecat STT Benchmark · semantic WER results table (pipecat-ai/stt-benchmark README) | 1 | 96% |
| prior · pull toward the middle | 4% | |
| weighted average | = 67.2 |
$ improve this page
Use a short GitHub form or edit the data through a pull request. Every number needs a source; here's which ones we can use.
data updated · sources and licences · json