Standard & Paid Price
Standard Price
| Meter | Standard Price |
|---|---|
| Input price | $0.1800 / audio hour |
Billing details
Description
StepFun StepAudio 2.5 ASR Stream — the streaming-recognition edition of StepAudio 2.5 ASR (a 4B-parameter recognizer built on Multi-Token Prediction, optimized for Chinese and English), and the streaming model StepFun recommends. On this gateway it is called like the other transcription models: upload a complete recording to /v1/audio/transcriptions and the full transcript is returned in one response once recognition finishes; partial results are not streamed back, and it costs more per audio minute than stepaudio-2.5-asr. Inverse text normalization is applied. Accepts 16-bit PCM wav and ogg uploads only (mp3 is not accepted); transcript output is not billed.
Capabilities
- Tasks
- transcription
API endpoints
POST /v1/audio/transcriptions(audio-transcription)