whatmodel.sh / stt / Ink-2
Ink-2
↗ share
+ contribute
Ink-2 is a closed-weights speech-to-text model from Cartesia, released May 2026. Licence: proprietary.
Catalogue identity notes
Cartesia's streaming STT model, API id ink-2. Introduced May 2026 English-only (Cartesia changelog, May 2026: 'Ink-2, our state-of-the-art streaming STT model ... It currently supports only English'); the alias was updated with a multilingual release on September 17, 2026 (docs model table: ink-2, release date September 17, 2026; changelog September 2026 'Ink-2 is now multilingual'). The Pipecat row was produced from the May English version (data added 2026-05-28, commit 5c1c6f6; TTFS figures unchanged at 9e24957 and 94f5672), run with language EN.
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 Ink-2, out of 176 with a score, then more from Cartesia.
| # | model | score | where | ||
|---|---|---|---|---|---|
| 5 |  Linden 1 (Agent STT) | 69±3 | api $0.0050/audio min | ||
| 6 |  Soniox v5 Async | 68±5 | api $0.0017/audio min | ||
| 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 | ||
| more from Cartesia | |||||
| 88 | Ink-Whisper | 55±3 | api price unknown | ||
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) | 1.0% ✎ | ▃·▃▅·▅▅·▃▅██ beats 11/16 | 96% |
$ where to run it
Through an API
api · $0.0065/audio min · docs.cartesia.ai ↗ · price sourceUSD per input audio minute: Cartesia's pricing page FAQ states 'Ink speech-to-text: One hour of audio is only $0.39 on the Scale plan. 3 credits equals 1 second of audio.' ($0.39/hr ÷ 60); the plan table lists Ink-2 as the STT model. Product: Cartesia Ink STT API (streaming). API model: ink-2. Settings: Scale plan ($299/month, 8M credits). Price basis: vendor-stated hourly cost on the Scale plan. Exclusions: plan fee, other plans' credit rates. Currency: USD. Cartesia 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 | = 67.9 |
$ 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