Transcription

POST /v1/audio/transcriptions takes multipart/form-data.

curl https://api.mixen.ai/v1/audio/transcriptions \
  -H "Authorization: Bearer $MIXEN_API_KEY" \
  -F file=@meeting.m4a \
  -F model=whisper-1 \
  -F language=en

Field	Description
file	Audio file, required
model	whisper-1 for compatibility
language	Language code; detected automatically when omitted
response_format	json by default

Billed per minute of audio (token-metered models bill per 1M tokens; see the catalog for units). The JSON response carries the charge: usage.cost / usage.cost_rub / usage.billing_source — same as every other protocol.

The live recognition line-up is in GET /v1/models, entries with type: stt:

Model	Key	What it is good for
MAI-Transcribe 2	microsoft/mai-transcribe-2	60 languages, strong multilingual accuracy — model card
GPT-4o Transcribe	openai/gpt-4o-transcribe	Familiar OpenAI quality
GPT-4o Mini Transcribe	openai/gpt-4o-mini-transcribe	Cheaper sibling, same interface
Voxtral Mini Transcribe	mistralai/voxtral-mini-transcribe	Mistral’s compact model
Nemotron 3.5 ASR Streaming	nvidia/nemotron-3.5-asr-streaming-multilingual-0.6b	The cheapest minute in the catalog
Scribe v2	elevenlabs/scribe-v2	90+ languages, word-level timestamps, audio event tagging
Gemini 3.5 Transcribe	google/gemini-3.5-transcribe	Word-level timestamps and diarization for up to 8 speakers, audio up to an hour

The endpoint requires the audio scope — a key without it gets a 403. The opposite direction — text to speech — is covered in the synthesis guide.