Context
262K
Input
$0.20/M
Output
$1.15/M
Aliases:step-3.7-flashstepfun/step-3.7-flashstep3.7-flashstep-37-flash
StepFun AI's flagship high-efficiency multimodal reasoning model on a sparse Mixture-of-Experts architecture (198B total, ~11B active) with a 256K context window, selectable reasoning effort, native image understanding, and multi-step function calling.
novita/step-3-7-flashQuantfp8
- Input
- $0.20/M
- Output
- $1.15/M
- Context
- 262K
- Max out
- 256K
- Cache read
- $0.04/M
- Cache write
- $0.20/M
- Thinking out
- $1.15/M