Vosk
Vosk is an offline speech recognition toolkit under Apache-2.0 built for small hardware. Its models are about 50 MB, it streams results with what its README calls zero-latency response, and it covers more than 20 languages with a reconfigurable vocabulary and speaker identification. Bindings exist for Python, Java, Node.js, C#, C++, Rust and Go, and it scales from a Raspberry Pi or an Android phone to a cluster.
What does Vosk do?
Vosk "is an offline open source speech recognition toolkit". Its README lists more than 20 languages and dialects, from English and Indian English through German, French, Spanish, Portuguese, Chinese, Russian, Turkish, Vietnamese, Italian, Dutch, Catalan, Arabic, Greek, Farsi, Filipino, Ukrainian, Kazakh, Swedish, Japanese, Esperanto, Hindi, Czech and Polish.
The design point is size and latency: "Vosk models are small (50 Mb) but provide continuous large vocabulary transcription, zero-latency response with streaming API, reconfigurable vocabulary and speaker identification." It scales "from small devices like Raspberry Pi or Android smartphone to big clusters".
Key facts
- Licence: Apache-2.0 on the toolkit. The licence of each model is not stated in the README.
- Languages listed: English, Indian English, German, French, Spanish, Portuguese, Chinese, Russian, Turkish, Vietnamese, Italian, Dutch, Catalan, Arabic, Greek, Farsi, Filipino, Ukrainian, Kazakh, Swedish, Japanese, Esperanto, Hindi, Czech and Polish.
- Models are about 50 MB and provide continuous large-vocabulary transcription.
- A streaming API with what the README calls zero-latency response, a reconfigurable vocabulary, and speaker identification.
- Bindings: Python, Java, Node.js, C#, C++, Rust, Go and others.
- Scales from a Raspberry Pi or an Android phone to large clusters, per the README.
- Installation, examples and documentation live on the project website rather than in the README.
- Latest tagged release when read: v0.3.50 on 2024-04-22, with repository commits in August 2026.
What does it replace, and where does it fall short?
Vosk replaces Deepgram and AssemblyAI for live recognition on constrained hardware, where a 50 MB model and a streaming API matter more than the last point of accuracy.
Where it falls short: the README is a pointer to the website, so word timestamps, speaker labelling beyond identification, quantisation, and the licence of each model are not stated here. The latest tagged release is from April 2024, though the repository had commits in August 2026.
How does Vosk run?
Through language bindings: Python, Java, Node.js, C#, C++, Rust, Go and others. Installation and examples are on the website. No Docker image, server or GPU requirement is described in the README.
Who is Vosk for?
A team putting recognition on a device, a kiosk or a board, or a chatbot that needs to react as the person speaks. A team wanting the highest accuracy on recordings should use Whisper or Parakeet.
What limits does the README state?
None listed.
Questions people ask
Is Vosk open source?
The toolkit is, under Apache-2.0. The README does not state the licence of the downloadable models, so check the model you use on the project website before shipping it. The code itself carries no use restrictions.
Does Vosk work offline?
Yes. The README's one-line description is "an offline open source speech recognition toolkit", and its models are small enough, about 50 MB, to sit on a phone or a Raspberry Pi with no server behind them.
Which languages does Vosk support?
More than 20, listed in the README: English and Indian English, German, French, Spanish, Portuguese, Chinese, Russian, Turkish, Vietnamese, Italian, Dutch, Catalan, Arabic, Greek, Farsi, Filipino, Ukrainian, Kazakh, Swedish, Japanese, Esperanto, Hindi, Czech and Polish.
How does Vosk compare with Deepgram?
Vosk gives you streaming recognition on hardware Deepgram's API cannot reach, a board or a phone with no network, with tiny models and a reconfigurable vocabulary. It does not claim the accuracy of the large models. The alternatives to Deepgram page places it as the pick for tiny models on boards and phones.
Is Vosk still maintained?
The repository had commits in August 2026 when read, though the last tagged release is v0.3.50 from April 2024. The weekly GitHub check on this site records both dates in the facts panel, so a change in either shows up here.
Sources
- Vosk README and Apache-2.0 LICENSE: github.com/alphacep/vosk-api, read 2026-09-16.
Compared with the others
On the open-source alternatives to Deepgram page, Vosk is the pick for tiny models on boards and phones. Apache-2.0. Models of about 50 MB, a streaming API, 20+ languages, speaker identification, and bindings from Python to Go, from a Raspberry Pi up to a cluster.
Also on that page: Whisper for the reference model on a gpu, whisper.cpp for a cpu, a mac, or a phone, WhisperX for word timestamps and speaker labels.
On the open-source alternatives to AssemblyAI page, Vosk is the pick for tiny models on boards and phones. Apache-2.0. Models of about 50 MB, a streaming API, 20+ languages, speaker identification, and bindings from Python to Go, from a Raspberry Pi up to a cluster.
Also on that page: WhisperX for word timestamps and speaker labels, Speaches for an openai-compatible server, Whisper for the reference model on a gpu.
More speech-to-text engines
OpenAI's general-purpose speech recognition model, released with code and weights under MIT, in six sizes from 39 million to 1.55 billion parameters, with multilingual transcription, translation to English and language detection.
A dependency-free C and C++ port of Whisper under MIT that runs on CPUs, Apple Silicon and phones, with quantised models, an HTTP server and bindings for a dozen languages.
Batched Whisper transcription under BSD-2-Clause with word-level timestamps from forced alignment and speaker labels from pyannote, the closest open-source match to a hosted transcription API's output.
NVIDIA's speech models and toolkit: Apache-2.0 code, and the Parakeet TDT 0.6B v3 model under CC-BY-4.0 with 25 European languages, automatic language detection, punctuation and word timestamps.
A runtime under Apache-2.0 from the next-generation Kaldi team that runs speech to text, text to speech, speaker diarization and voice activity detection locally on CPUs, NPUs and phones, with no internet connection.
An on-device speech toolkit under MIT built for live streaming, with speech-to-text models trained from scratch in sizes down to 1 MB, one library across Python, the browser, phones and desktops.
Added September 16, 2026. Every claim above comes from the project's README, LICENSE or model card, read on September 16, 2026, or from the GitHub API on the date shown in the panel. Found an error? Write to reveneau@licheo.com and it is fixed in the next weekly pass. Repository: github.com/alphacep/vosk-api.
Want Vosk running on your own servers?
Reveneau sets it up on your hardware, connects it to your product, and builds the software around it.