JavaScript speech-to-text API

The speech-to-text API for JavaScript developers

Transcribe audio files or live microphone streams from Node, TypeScript, or the browser. One typed SDK returns accurate text with timestamps and speaker labels—no Web Speech API quirks. Start free, no credit card.

Start building free

support-call.wav Done · 0:38

Speaker A 0:03

Thanks for calling—can you confirm the order number on the invoice?

Speaker B 0:09

Sure, it's the shipment that went out to Portland last Tuesday.

Speaker A 0:16

Found it. I'm issuing the replacement now—you'll get a confirmation email.

TypeScript types Real-time WebSocket Speaker labels

Typed end to end

npm install assemblyai gives you a TypeScript-first SDK—autocomplete on every field of the transcript, from utterances to entities.

Real-time from the mic

Stream audio over a WebSocket and partial and final transcripts come back within a few hundred milliseconds—in every browser, not just Chrome.

Audio files to text

Await one transcribe call with a file or URL and the SDK handles upload and polling, returning formatted text with timestamps.

36%

improvement in close rate

See case study

“The new Universal-3.5 Pro speech model from AssemblyAI is best so far in terms of accuracy, latency, and language switching.”

80%

increase in customer satisfaction

“Assembly has saved us countless hours managing models, and provided exceptional accuracy.”

36%

improvement in close rate

See case study

“The new Universal-3.5 Pro speech model from AssemblyAI is best so far in terms of accuracy, latency, and language switching.”

80%

increase in customer satisfaction

“Assembly has saved us countless hours managing models, and provided exceptional accuracy.”

Browser vs server

More reliable than the Web Speech API

The browser's built-in SpeechRecognition is free, but it's Chromium-biased, inconsistent across browsers, and gives you no control over the model or your users' audio. A speech-to-text API runs server-side: the same accuracy in Chrome, Safari, and Firefox, plus speaker labels, timestamps, and audio that stays under your data agreement.

Start building free

No credit card required

Streaming word error rate

Word error rate is the share of words the model gets wrong against a human reference—the standard measure of accuracy for live, streaming transcription.

Streaming word error rate on English audio.

*Lower is better*

Streaming word error rate (English)

AssemblyAI Universal-3.5 Pro Realtime

5.53%

Deepgram Flux

8.87%

Deepgram Nova-3

9.39%

Cartesia

10.10%

Source: AssemblyAI published benchmarks — assemblyai.com/benchmarks.

Common questions

: Run npm install assemblyai, create a client with your free API key, and await a single transcribe call with a file or URL. For live audio, open a WebSocket stream with the same SDK and transcripts arrive as your users speak.
: For live audio, compare streaming accuracy: on published benchmarks, AssemblyAI's Universal-3.5 Pro Realtime has the lowest streaming word error rate (5.53%) against Deepgram Flux, Deepgram Nova-3, and Cartesia. See the full benchmarks, then test the API on your own audio—that's the comparison that matters.
: Yes. Capture mic audio the way you already do and stream it to a speech-to-text API over a WebSocket. You get consistent results in Chrome, Safari, and Firefox—the Web Speech API's biggest limitation—plus control over the model and where your users' audio is processed.
: Yes. The JavaScript SDK is built for Node and TypeScript—transcribe files in a batch job, stream live audio server-side, or power a voice agent, all with typed, promise-based calls.
: Yes. Call the API from your server routes, or stream directly from the browser using short-lived tokens so your API key never ships to the client. The SDK works anywhere JavaScript runs.
: AssemblyAI's free tier includes up to 185 hours of pre-recorded transcription and up to 333 hours of streaming—no credit card required. After that, pay-as-you-go pricing applies with no upfront commitments.