Step 3.7 Flash is StepFun's latest high-efficiency multimodal Mixture-of-Experts model. It pairs a 196B-parameter language backbone with a vision encoder for native image and video understanding, activating roughly 11B parameters...
Rank
#15
Down 18% week over week
Tokens
1.06T
2026-08-14
Requests
14.9M
14,865,985
Tool calls
-
Input / M
$0.20
Output / M
$1.15
Cache / M
$0.04
Context
262K
262,144 tokens