AssemblyAI Docs

Appearance

AssemblyAI DocsVoice AI Infrastructure

Production-ready speech-to-text, audio intelligence, and LLM Gateway. Build with pre-recorded or streaming audio using REST APIs and WebSockets.

Get Started
API Reference
🎤

Pre-Recorded Speech-to-Text

High-accuracy transcription with Universal-3 Pro and Universal-2. Supports 100+ languages, speaker diarization, medical mode, and custom prompting.

📡

Streaming STT

Real-time transcription with sub-300ms latency. Universal-3 Pro Streaming with turn detection, barge-in support, and partial transcripts.

🤖

Voice Agent API

Build real-time voice AI agents with a single WebSocket connection. Speech in, speech out. Full tool calling and browser integration support.

🧠

LLM Gateway

Unified API for 25+ models from Claude, GPT, Gemini, and more. Apply LLMs directly to your audio transcripts for summaries, Q&A, and analysis.

🔍

Speech Understanding

Extract structured insights from audio: sentiment analysis, entity detection, speaker identification, topic detection, key phrases, and more.

🛡️

Guardrails

PII redaction, content moderation, profanity filtering, and speech thresholds to protect sensitive content in your applications.

Quick Start ​

python
import assemblyai as aai

aai.settings.api_key = "YOUR_API_KEY"

transcriber = aai.Transcriber()
transcript = transcriber.transcribe("https://example.com/audio.mp3")

print(transcript.text)
typescript
import { AssemblyAI } from 'assemblyai'

const client = new AssemblyAI({ apiKey: 'YOUR_API_KEY' })

const transcript = await client.transcripts.transcribe({
  audio: 'https://example.com/audio.mp3'
})

console.log(transcript.text)
bash
curl -X POST https://api.assemblyai.com/v2/transcript \
  -H "Authorization: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"audio_url": "https://example.com/audio.mp3"}'

Available Models ​

ModelUse CaseLatencyLanguages
Universal-3 Pro (Async)Highest accuracy, pre-recorded15-30% of audio duration6 languages
Universal-2Broad language support15-30% of audio duration99+ languages
Universal-3 Pro StreamingReal-time, voice agentsUnder 300ms6 languages
Universal StreamingReal-time, multilingualUnder 500ms60+ languages
Whisper Streaming99+ language streamingVaries99+ languages
  • Meeting Notetaker: Transcribe + diarize + LLM Gateway for structured notes
  • Contact Center Analytics: Sentiment, key phrases, speaker roles, coaching
  • Medical Scribe: Medical Mode + SOAP note generation via LLM Gateway
  • Voice Agent: Real-time STT + TTS with interruption handling
  • Live Captioner: Streaming STT with keyterms for domain accuracy