Pass IndexThe State of AISign in

stepaudio-2.5-asr

StepFun's next-generation flagship ASR model, a 4B-parameter model with a Multi-Token Prediction architecture that can transcribe five minutes of audio in under a second.

audio → text · made by StepFun

$0.00037per minute
StepFun

Sold by 1 way

SellerLaneRate
StepFun apistandard$0.00037per minuteplatform.stepfun.com · read 2026-08-25

About

stepaudio-2.5-asr — a text model from StepFun, sold by one company from $0.00037 per minute.

It takes audio and returns text. It was published in April 2026. The catalogue files it under transcribe. Only StepFun sells it, at $0.00037 per minute.

Every current figure

Maker
StepFun
Register
model
Takes
audio
Returns
text
Published
April 2026
Licence
not read
Sellers
1
Price
$0.00037 per minute — StepFun