keirouter-stt

KeiRouter — Speech-to-Text

Requires KEIROUTER_URL (and KEIROUTER_KEY if auth enabled). See https://raw.githubusercontent.com/mydisha/keirouter/main/skills/keirouter/SKILL.md for setup.

Discover

curl $KEIROUTER_URL/v1/models/stt | jq '.data[].id'
# Per-model params (language, response_format, prompt, temperature support)
curl "$KEIROUTER_URL/v1/models/info?id=openai/whisper-1"

model = STT model ID (e.g. openai/whisper-1, groq/whisper-large-v3, deepgram/nova-3, gemini/gemini-2.5-flash).

Endpoint

POST $KEIROUTER_URL/v1/audio/transcriptions (OpenAI Whisper compatible, multipart/form-data)

Field	Required	Notes
model	yes	from /v1/models/stt
file	yes	audio file (mp3, wav, m4a, webm, ogg, flac)
language	no	ISO-639-1 (e.g. en, vi)
prompt	no	hint text to guide transcription
response_format	no	json (default) / text / verbose_json / srt / vtt
temperature	no	0–1

Examples

curl -X POST "$KEIROUTER_URL/v1/audio/transcriptions" \
  -H "Authorization: Bearer $KEIROUTER_KEY" \
  -F "model=openai/whisper-1" \
  -F "[email protected]" \
  -F "language=en"

JS (Node):

import { createReadStream } from "node:fs";
const form = new FormData();
form.append("model", "groq/whisper-large-v3-turbo");
form.append("file", new Blob([await (await import("node:fs/promises")).readFile("audio.mp3")]), "audio.mp3");
const r = await fetch(`${process.env.KEIROUTER_URL}/v1/audio/transcriptions`, {
  method: "POST",
  headers: { "Authorization": `Bearer ${process.env.KEIROUTER_KEY}` },
  body: form,
});
const { text } = await r.json();
console.log(text);

Response shape

Default (response_format=json):

{ "text": "Hello, this is the transcription." }

verbose_json adds language, duration, segments[] with timestamps. srt / vtt return subtitle text.

Provider quick reference

Provider	model format	Notes
OpenAI	whisper-1, gpt-4o-transcribe, gpt-4o-mini-transcribe	Native OpenAI shape
Groq	whisper-large-v3, whisper-large-v3-turbo	Fastest; OpenAI shape
Gemini	gemini-2.5-flash, gemini-2.5-pro	Server converts to generateContent with audio
Deepgram	nova-3, nova-2	Token auth; server adapts response
AssemblyAI	universal-3-pro, universal-2	Async upload+poll handled server-side