

Wispr · od 17.09.2026 (öffentliche Vorstellung; Vorschau am 17.08.2026) · 6× · naposledy 11. 11. 2026
50
Momentum
Canto is Wispr's first proprietary speech model (about 2 billion parameters), built for real-time dictation and used inside the Wispr Flow app. It was previewed on August 17, 2026 and publicly detailed on September 17, 2026. Wispr reports the lowest word error rate in its own tests; these vendor tests are not independently verified.
Vývoj momenta
18.08.16.11.
Vlastnosti
| Real-Time Streaming | Yes, designed for real-time dictation |
| Platform | Cloud transcription, integrated in Wispr Flow (macOS, Windows, iOS, Android) |
| Price | No separate Canto price; via Wispr Flow: Free 2,000 words/week (desktop), Pro $15/month ($12 billed annually) |
| Release Date | September 17, 2026 (publicly introduced) |
| Languages | Launch evaluation English only; Wispr Flow overall 100+ languages with auto-detection |