Vosk
An offline speech-recognition toolkit with models for more than twenty languages.
What’s included for free?
Open-source local or self-hosted software with no provider-enforced usage quota; hardware, models, hosting or external APIs may be separate.
Provider information ↗Checked 2026-10-23
Before you use it
Apache-2.0; model notices may differ.
Provider terms can change. Check current pricing and usage rights before starting a project.