customers
All customer stories
Top Voice AI companies are building with Assembly.
resources
Latest Release
Text-to-Speech API
Our first text-to-speech model, AssemblyAI-TTS-1, is built for voice agents, on live calls. It sounds natural, responds fast, and keeps callers on the line.
resources
The Voice AI infrastructure for every workflow
Production Voice AI from a single API: models, intelligence, deployment.
How does your audio arrive?
Your audio type determines the right product.
Realtime Speech-to-Text API
Transcribe live audio and video files in real time at ultra-low latency and high-quality accuracy.
Text-to-Speech API
Trained on real conversational speech, so your voice agents sound human, get names and numbers right, and keep callers on the line.
Medical Mode
Optimize transcription for medical terminology and healthcare conversations with significantly improved accuracy. Available on Universal-3.6 Pro Realtime and Universal-Streaming models. BAA available.
Dictation API
Introducing the first API purpose-built for dictation. Text comes back already formatted by combining speech-to-text with an LLM cleanup pass in a single call.
Sync Speech-to-Text API
Finished transcripts in a single API call. Send a short clip and read the transcript off the response — no polling, no WebSocket, no job to manage.
Turn transcripts into structured intelligence
Add modular intelligence and safety layers on top of any transcript.
Speech Understanding API
Extract structured insights from speech without building custom NLP pipelines.
LLM Gateway
Send transcripts directly to GPT, Claude, Gemini, or open-source models via a single API.
Guardrails
Compliance-grade safety controls at the transcription layer.
Voice AI infrastructure that scales with you
Run on AssemblyAI's managed cloud or deploy on your own infrastructure. Same models, same API.
Purpose-built for the hardest Voice AI problems
The same API powers voice agents, clinical documentation, meeting notes, and contact centers at scale.
Voice Agents
Entity-accurate real-time transcription with turn detection and short-utterance handling — the model stack that wins competitive voice agent evals.
AI Notetakers
Highest accuracy with speaker diarization, custom output formatting via prompting, and LLM Gateway for automatic summaries, chapters, and action items.
AI Scribe
Ambient clinical documentation powered by Medical Mode — ~20% reduction in missed entities on drug names, conditions, and procedures. HIPAA BAA available in minutes.
Conversation Intelligence
Turn every customer conversation into structured data — sentiment analysis, entity detection, topic classification, and key phrases extracted automatically from transcripts.
Agent Assist
Real-time streaming transcription that powers live agent coaching, suggested responses, and compliance monitoring during active customer calls.
Call Analytics
Post-call transcription with speaker diarization, sentiment tracking, and LLM-powered QA scoring. Process call recordings at scale for trends, compliance, and coaching insights.