tools
Flow
Flow is Speechmatics' voice AI agent for natural spoken conversations with real-time transcription and spoken responses.
Flow listens to microphone audio, transcribes speech with Speechmatics technology, sends the transcript to third-party large language models, and speaks the response using text-to-speech. It is designed for voice interactions that can handle interruptions, crosstalk, accents, dialects, and multiple speakers.
Flow is available through an exclusive testing program and is provided free for experimental and entertainment purposes. It is not for commercial exploitation, consequential decisions, or reliance on its outputs. Speechmatics' pricing page lists separate Speech API plans, but no Flow-specific paid plan.
Features
- Transcribes microphone audio in real time
- Generates responses through third-party large language models
- Speaks responses using text-to-speech
- Handles interruptions and crosstalk
- Supports speaker identification and diarization
- Designed for different languages, accents, and dialects
- Hosted on Speechmatics infrastructure
- Free for experimental and entertainment purposes
Use cases
- Test natural voice conversations with an AI agent
- Prototype voice assistants and spoken interfaces
- Demonstrate interruption-aware voice interactions
- Explore multi-speaker voice AI behavior
- Evaluate speech recognition with accents and background noise
Pros
Cons
Pricing
- Starting price
- Free
- Pricing checked
- 2026-09-19
Free
$100 free
- 55+ languages
- $100 in credit to get started
- 2 concurrent real-time sessions
- Multi-region cloud options
- Low-latency text-to-speech
Pro
from$0.129/hr
- 55+ languages
- $100 in credit to get started
- 50 concurrent real-time sessions
- 10 file jobs per second
- Multi-region cloud options
- Online email support
Enterprise
Get in touch
- Unlimited scale
- Flexible deployments
- No rate limits
- Privacy-first deployment options
- Custom models
- Prioritized service and support