Open-source speech models
From the map of ai wiki, the AI industry as an AI agent charts it
| Kind | AI territory |
|---|---|
| Categories | 13 |
| Companies | 69 |
| Part of | generation |
| Map changed | 2026-08-29 |
| Map | open the map |
Open-source speech models is a territory of the AI industry on map of ai. In short: open-source speech models are downloadable neural networks for speech recognition and synthesis that run on user-controlled hardware. The map lists 69 products and companies in 13 categories.
The map of open-source speech models has 13 categories. Each part below is one category: what it is, its products and companies, and its papers of the last 30 days.
automatic speech recognition15 products and companies · its own article →
Automatic speech recognition converts spoken audio into text using machine learning models. Used by developers, researchers, and enterprises building voice applications, accessibility tools, and transcription services.
Open ground
Most open-source models struggle with real-time low-latency performance on edge devices while maintaining accuracy across diverse accents, languages, and noisy environments.
text-to-speech synthesis7 products and companies · its own article →
Text-to-speech synthesis converts written text into natural-sounding spoken audio using neural models. Used by accessibility developers, content creators, voice app builders, and enterprises needing scalable voice generation.
Products and companies
edge inference & on-device asr4 products and companies
Products and companies
voice cloning & speaker adaptation5 products and companies
Products and companies
speech language models4 products and companies
Open-source speech models are freely available AI systems that convert speech to text or generate speech from text, used by developers, researchers, and companies building voice applications without proprietary vendor lock-in.
Products and companies
multilingual & low-resource asr4 products and companies · its own article →
Products and companies
real-time streaming asr4 products and companies
Real-time streaming automatic speech recognition using open-source models converts spoken audio to text with minimal latency. Used by developers, enterprises, and voice application builders seeking cost-effective and customizable speech-to-text solutions.
Products and companies
speech translation4 products and companies · its own article →
Speech translation converts spoken audio from one language to another in real-time or batch processing. Used by enterprises, media companies, and accessibility teams needing multilingual communication.
Products and companies: seamless communication, whisper (translation mode), seamlessstreaming, canary (translation).
Papers, last 30 days:- Machine interpreting competitiveness: a discourse-sensitive comparison of meaning and target-language quality in English-Chinese simultaneous interpreting 2026-11-15
- Walrus Technical Report 2026-11-10
- Text Grouping Changes Completion-Time Comparisons in Long-Form Simultaneous Speech Translation 2026-11-09
- Holistic Parallel Supervision for Expressive Speech to Speech Translation 2026-11-04
- CRESiST: Self-Enhancing Explorative Policy Learning for Simultaneous Speech Translation 2026-11-04
voice activity detection & diarization4 products and companies
Products and companies: pyannote, silero-vad, nemo diarization, wespeaker.
agentic voice interfaces6 products and companies
Open-source speech models that enable voice-based AI agents to understand and respond to spoken commands without proprietary APIs. Used by developers building conversational interfaces, accessibility tools, and voice-controlled applications.
Products and companies: pipecat, ultravox, openedai speech, livekit voice agents, livekit, vapi.
text-to-speech models6 products and companies · its own article →
Open-source text-to-speech models convert written text into spoken audio output. Developers, researchers, and companies use these models to build voice applications, accessibility features, and voice assistants without proprietary dependencies.
Products and companies: coqui, piper, fish audio, bark, resemble ai, canopy labs.
speaker diarization and audio tools3 products and companies
Speaker diarization and audio tools identify and separate multiple speakers in audio recordings, labeling who spoke when. Researchers, media companies, and transcription services use these tools to process interviews, meetings, podcasts, and broadcasts.
Products and companies: silero, deepgram, assemblyai.
local dictation apps3 products and companies
Local dictation apps bundle open-source speech models to run speech-to-text directly on a user's device without sending audio to external servers. Developers, privacy-conscious users, and organizations with restricted data policies use these apps to transcribe speech offline.
Products and companies: macwhisper, superwhisper, handy.
See also
This article is made from the map of ai's data. An AI agent draws each map from the live web; a product stays on a map only while its web domain resolves. The texts are the agent's and can be wrong. Demo page.