Boson AI
Real-time voice agents and audio models for business workflows
About Boson AI
Boson AI provides foundational audio models and voice-agent capabilities for business workflows, enabling agents to learn from every call. Higgs Realtime supports real-time voice interaction, speech-to-speech, tool use, and low-latency responses for support hotlines, sales calls, and product assistants. Higgs Audio / Avatar covers text-to-speech, speech-to-text, and avatar models, with production workflows trained for business domains on business data.
Best for
Best use cases
Support hotlines
Sales calls
Product assistants
Business-critical workflows
Features
Real-time voice agent
Supports real-time listening, reasoning, tool use, and responses with an emphasis on low-latency interaction.
Voice and audio models
Covers speech-to-speech, text-to-speech, and speech-to-text capabilities.
Avatar models
Provides avatar visuals for natural voices, giving sound a visual presence.
Business workflow adaptation
Can be trained for business domains on business data and run in production.
Pros
- Supports real-time listening, reasoning, tool use, and responses
- Can be trained for business domains on business data
Pricing Plans
Free trial credit
- Free trial credit for new accounts
Text to speech
- Generated audio
Input tokens
- Input tokens
Cached input tokens
- Cached input tokens
Output tokens
- Output tokens
Transcription input
- Transcription input
* Pricing information is for reference only. Please refer to the latest announcement on Boson AI's official website.