Vosk provides an offline speech recognition API for mobile devices and servers, with several programming language interfaces.
Open source
24 alternativesKaldi
Kaldi provides an open toolkit for speech recognition and speech processing research.
About Kaldi
Kaldi is a toolkit for speech recognition and speech processing research. The official project location is a GitHub repository, and its listed topics include speech-to-text and speaker verification. The codebase is C++ and is built from source using CMake.
The repository includes build instructions for UNIX-like systems, Darwin, and Cygwin, with separate Windows instructions. It also provides example system builds and platform notes for Fedora 41, ppc64le, Android, and WebAssembly. The README warns that users will probably encounter problems and directs them to the developers and to mailing lists for help. Pricing is not stated.
Key features
- Speech recognition and speech-to-text toolkit
- Speaker verification and speaker identification topics
- Example system builds documented in the egs directory
- Separate build instructions for UNIX-like systems and Windows
- Cross-compilation for Android using Android NDK and OpenBLAS
- In-browser execution via WebAssembly using emscripten and OpenBLAS
Best for: Researchers and developers working on speech recognition or speaker verification who are comfortable building software from source.
Good for
- Covers speech recognition and speaker verification in one toolkit
- Includes build notes for several platforms, including Android and WebAssembly
- Provides example system builds to run as a starting point
Watch out for
- Requires building from source, with different instructions for each platform
- Some variants are untested, such as Cygwin on less common UNIX systems
- The README states that users will probably encounter problems and points them to developer contact channels
Kaldi at a glance
- Website
- github.com/kaldi-asr/kaldi
- Category
- AI & Machine Learning › AI Voice & Audio
- Type
- Open-source project
- Pricing
- Not confirmed. Check the site
- Platforms
- Linux, Windows
- Open source
- Not stated
- Language
- English
- Good for
- Developers
- Last checked
- Oct 6, 2026 · Online when checked
- Listed
- Oct 5, 2026
Kaldi alternatives
Tools and sites you could use instead, ranked by how closely they match.
The wav2letter project provides an automatic speech recognition toolkit from Facebook AI Research.
ESPnet includes tools for end to end speech processing and related speech tasks.
Open source
24 alternatives
The whisper.cpp project ports OpenAI’s Whisper speech recognition model to C and C++.
Open source
24 alternatives
Superwhisper turns speech into polished text for use in apps such as Slack and Gmail, with recognition options listed for four operating systems.
Offer Bull provides AI interview practice, real-time speech recognition, and live answer suggestions for job seekers preparing for online interviews.
Popular in AI Voice & Audio
Well-known listings from the same topic.
Google Illuminateilluminate.google.comAlteredaltered.aiELSA Speakelsaspeak.comKreadoAIkreadoai.comPodwisepodwise.aiAddsubtitleaddsubtitle.ai
Kaldi FAQ
How can contributors submit changes?
Create a personal fork of the main repository, make changes in a named branch other than master, and open a pull request through the GitHub web interface. The project generally follows the Google C++ Style Guide, with some exceptions.
Where can I find example system builds?
The README says to see egs/README.txt in the repository for instructions on running the example system builds.
Is something wrong or out of date?
Tell us and an editor will check it.
Own this website?
Claim this listing
Claim the listing and suggest updates yourself. An editor checks every claim first.