Standard & Paid Price
Standard Price
| Meter | Standard Price |
|---|---|
| Input price | $1.4815 / 1M tokens |
| Completion price | $10.3704 / 1M tokens |
| Cache read price | $0.2963 / 1M tokens |
Billing details
Description
StepFun step-audio-2 — end-to-end speech model that consumes audio directly rather than working from a transcript, so tone and delivery survive into the model's understanding. Billed per token, sharing the input rate of the StepAudio 2.5 line at a higher output rate. StepFun publishes no separate capability page for this model; consult the platform audio documentation for the current parameter set.
Capabilities
- Tasks
- chat, audio-understanding
API endpoints
POST /v1/chat/completions(openai)