Deepgram

Provides a voice AI platform with speech-to-text, text-to-speech, voice agent, and audio intelligence APIs that converts audio to text, generates spoken audio, and enables conversational voice agents. Sells to developers and product teams at startups and enterprises, plus platforms and partners embedding voice capabilities and enterprises needing custom models. Unifies speech-to-text, text-to-speech, and LLM orchestration into a single voice agent API instead of stitching separate components, available in real time or batch in the cloud or self-hosted.

Key features

• Real-time speech-to-text transcription
• Batch speech-to-text transcription
• Text-to-speech with expressive voices
• Voice Agent API single call
• LLM orchestration for voice agents
• Audio Intelligence from conversations
• Conversation-aware turn-taking and interruptions
• Cloud or self-hosted deployment
• Real-time or batch processing