Speech to text with word timestamps

Transcription

Transcribe audio into text with word-level timestamps through the audio transcriptions API.

Get startedRead the docs
POST /v1/audio/transcriptions
3.1s
0.26sHello,1.08swelcome1.58sto2.03sLLM2.65sGateway.
grok-stt-1-0 · 3.49s of audio
01

Word timestamps

Start and end time per word.

02

Language detection

Returned with every result.

03

Low cost

Billed per hour of audio.

04

One API key

Same key as your chat calls.

Built for

  • Captions

    Subtitles synced to the word

  • Call notes

    Transcripts for meetings and calls

  • Voice input

    Turn speech into prompts

Transcribe a file
terminal
1curl https://api.llmgateway.io/v1/audio/transcriptions \2  -H "Authorization: Bearer $LLM_GATEWAY_API_KEY" \3  -F model=grok-stt-1-0 \4  -F file=@call.mp3
Read docs

More in models & modalities

All features

Embeddings

Vectors for search and RAG

Rerank

Sharper retrieval in one call

Search API

Ranked web results with page content

Start using Transcription today

One API key. Every model. Live in minutes.

Start buildingBrowse models