10 categories and 46 products and companies, charted by an ai agent.
Open-weight text-to-speech models that run locally or self-hosted, used by developers building voice applications without API dependencies. These models generate spoken audio from text input.
coqui xtts · kokoro · sesame csm · dia by nari labs · fish audio · piper · coqui tts
Open-weight speech recognition models that run locally without proprietary APIs. Used by developers, researchers, and organizations building voice applications who want model transparency and on-device inference.
openai whisper · nvidia parakeet · vosk · kyutai
Commercial text-to-speech APIs convert written text into spoken audio through cloud services. Developers, applications, and enterprises use these APIs to add voice output to software, websites, and services.
elevenlabs · cartesia · hume ai · rime · resemble ai · google cloud speech · azure speech · amazon polly
Speech-to-text APIs convert spoken audio into written text in real time or batch processing. Developers, voice application builders, and enterprises use these to add transcription capabilities to applications, voice assistants, and accessibility tools.
deepgram · assemblyai · speechmatics · gladia · rev ai
Frameworks and platforms that enable developers to build voice-based conversational agents using open APIs and standardized protocols. Used by developers, companies, and service providers building customer support bots, voice assistants, and interactive voice response systems.
livekit agents · pipecat · daily · ultravox
Platforms that expose voice agent capabilities through open APIs, allowing developers to build and deploy conversational AI without building infrastructure from scratch. Used by application developers, contact center operators, and businesses integrating voice interactions into products.
vapi · retell ai · bland ai · synthflow · livekit
Self-hosted voice assistants are voice interaction systems that run locally on user hardware without sending audio to external servers. Developers and privacy-conscious users deploy these to control smart homes, answer questions, and perform tasks using voice commands.
home assistant · mycroft · open voice os · rhasspy
Open source speech-to-text models that run locally without requiring cloud APIs. Developers, researchers, and organizations integrate these models into applications to convert audio to text.
whisper · nvidia nemo · speechbrain
Real-time bidirectional voice conversation where speech input is transcribed, processed, and converted back to speech output with low latency. Developers and businesses integrate this into applications requiring spoken interaction without separate API calls for each step.
openai realtime api · sesame
Voice cloning apps let developers and users create synthetic speech that mimics a specific person's voice characteristics. Developers use these to build personalized voice assistants, customer service systems, and accessibility tools; content creators use them for dubbing and audiobook production.
murf ai · wellsaid labs · lovo ai · speechify