AsyncFlow API
  1. Integrations
AsyncFlow API
  • Welcome to AsyncFlow API
  • Get Started
  • API Reference
    • API Status
      • API Status Check
    • Text-to-Speech
      • Text to Speech (WebSocket)
      • Text to Speech
      • Text to Speech with Word Timestamps
      • Text to Speech (Stream)
    • Voice Management
      • Clone Voice
      • List Voices
      • Get Voice
      • Update Voice
      • Delete Voice
      • Get Voice Preview
  • Advanced Guides
    • Custom Pronunciations
      • Embedding Custom Phonemes in Async TTS API
      • Pronouncing digits one‑by‑one
  • Integrations
    • Integrate with Twilio
    • Pipecat Integration
  1. Integrations

Pipecat Integration

Async delivers low‑latency, high‑fidelity text‑to‑speech (TTS) that plugs into Pipecat through our official integration.
Pipecat exposes two Async service classes:
Service classTransportStreaming
AsyncAITTSServiceWebSocket (persistent)Low-latency chunks
AsyncAIHttpTTSServiceHTTP (per‑request)Response streaming
AsyncAITTSService is recommended for interactive use because it streams audio chunks as soon as they are ready, minimising time‑to‑first‑byte (TTFB).

Prerequisites#

Python 3.10 +
Pipecat 0.0.77 or later
Async account & API key
(Optional) preferred Async voice ID

Installation#

Set your API key once:

Quick‑start (WebSocket)#

Handling interruptions#

AsyncAITTSService automatically stops synthesis on user interruption or when it receives a TTSUpdateSettingsFrame, enabling natural, back‑and‑forth voice experiences.

Troubleshooting#

SymptomLikely causeFix
ValueError: Missing ASYNCAI_API_KEYEnv var not exportedexport ASYNCAI_API_KEY=...
Long TTFB (>1 s)Using HTTP service / cold startSwitch to WebSocket & keep connection warm
ErrorFrame quota exceededCharacter quota exhaustedUpgrade plan

Next steps#

📖 Read the Pipecat reference
🛠 Browse example code

Modified at 2025-08-01 13:07:39
Previous
Integrate with Twilio
Built with