AssemblyAI is a speech AI company founded in 2017 by CEO Dylan Fox. The company builds APIs that turn audio and voice into structured data and conversational experiences, spanning pre-recorded and real-time transcription, fully managed voice agents, and audio intelligence add-ons like sentiment analysis and PII redaction. AssemblyAI has raised approximately $145M across three funding rounds, most recently a $50M Series C in December 2023 led by Accel.
| General | |
|---|---|
| Company | AssemblyAI |
| Founded | 2017 by Dylan Fox |
| Headquarters | San Francisco, California, United States |
| Website | assemblyai.com |
| Documentation | assemblyai.com/docs |
| GitHub | github.com/AssemblyAI |
| Type | Speech AI / Voice AI infrastructure |
Core Products
Speech-to-Text
Transcribes pre-recorded audio and video across 99 languages using the Universal-3.5 Pro model, with speaker diarization, PII redaction, and native code-switching for audio that moves between languages.
Universal-Streaming
Real-time transcription over WebSocket for live captioning and voice agents, built on the Universal-3.5 Pro Realtime model with sub-second latency.
Voice Agent API
A fully managed speech-to-speech pipeline combining transcription, LLM reasoning, text-to-speech, turn detection, and tool calling behind a single WebSocket connection.
Speech Understanding API
Layers structured intelligence on top of transcripts: sentiment analysis, entity detection, topic detection, summarization, translation, and speaker identification.
Guardrails API
Safety and compliance controls for audio and transcripts, including PII redaction, profanity filtering, and content moderation.
LLM Gateway
An OpenAI-compatible API for calling frontier language models like Claude, GPT, and Gemini inside voice AI workflows, replacing the deprecated LeMUR product.
Developer Resources
AssemblyAI provides official SDKs for Python and JavaScript/TypeScript, starter templates for voice agents, and $50 in free credit at signup with no credit card required.
Helpful Links
- Documentation: API reference, guides, and quickstarts
- GitHub: official SDKs and starter templates
- Pricing: per-product rates and free tier details
- Discord: community support
Key Features
99-language transcription with native code-switching Universal-3.5 Pro transcribes 99 languages and natively handles audio that switches between 18 of them mid-conversation, without requiring a language to be selected in advance.
Fully managed voice agent pipeline The Voice Agent API bundles speech-to-text, LLM reasoning, text-to-speech, turn detection, and tool calling into one WebSocket connection, with native plugins for LiveKit and Pipecat.
Modular audio intelligence Sentiment analysis, entity detection, topic detection, PII redaction, and content moderation are available as add-ons on top of any transcription request, priced and billed independently.
Use Cases
Voice agents and conversational AI Build speech-to-speech agents with the Voice Agent API, deployed through LiveKit, Pipecat, Twilio, or Telnyx telephony.
Contact center and sales call analytics Combine transcription with sentiment analysis, topic detection, and entity detection to analyze calls at scale.
Meeting intelligence and medical documentation Use contextual prompting and Speech Understanding add-ons to generate accurate transcripts and summaries for meetings and clinical audio.
.png&w=96&q=75){kind=small}
.png&w=96&q=75){kind=small}
.png&w=1920&q=75)
