Every open model that has done speech to text in English on the network, fastest first. Speed is how many times faster than real time the machine produced it; below 1× it is slower than real time.
| Model | Family | Speed | Machines |
|---|---|---|---|
| moonshine tiny | Moonshine | 41× real time | 6 |
| Whisper large-v3 | Whisper | 5.1× real time | 1 |
| moonshine base | Moonshine | 4.7× real time | 5 |
| Whisper small | Whisper | 0.4× real time | 3 |
| Seamless M4T v2 Large (speech → text) | Seamless M4T v2 | 0.2× real time | 3 |
| Whisper large-v3-turbo | Whisper | 0.1× real time | 6 |
| Whisper medium | Whisper | 0.1× real time | 2 |
| MiniCPM-o 4.5 BASE | MiniCPM-o 4.5 | measured | 2 |