Model creator
StepFun
Every StepFun model in the BenchSift catalogue, with its scores, ranks, prices and capabilities as the sources publish them.
- Models
- 13
- Open weights
- 5
- 5 of 13 models
- Latest release
- Sep 18, 2026
- Step 5 Preview · First dated model: Jan 20, 2026
- Median price
- $0.294
- $0.15 to $1.43 per million tokens, across 4 priced models
Models
13Reasoning levels are folded into one entry. The rank is the whole catalogue's: sorting does not change it.
- Step 5 Preview28.5#145 of 782$1.43149 t/s
Text · Reasoning · Open weights · Sep 18, 2026 · $1.43 · 149 t/s
- Step 3.7 Flash19.5#273 of 782$0.438144 t/s
Text · Reasoning · Open weights · May 29, 2026 · $0.438 · 144 t/s
- Step 3.5 Flash 260317.0#303 of 782$0.15102 t/s
Text · Reasoning · Open weights · Apr 2, 2026 · $0.15 · 102 t/s
- Step 3.5 Flash17.0#303 of 782$0.15125 t/s
Text · Reasoning · Open weights · Feb 2, 2026 · $0.15 · 125 t/s
- Step3 VL 10B7.7#541 of 782——
Text · Reasoning · Open weights · Jan 20, 2026
- ————
- Step Image Edit 2————
Image
- ————
- Step-Video-T2V————
Video
- Step1X-Edit————
Image
- Step1X-Edit-v1p2————
Image
- ————
- ————
Where these figures come from
Each catalogue model is tied to its creator by the identity Artificial Analysis publishes, or OpenRouter's and Models.dev's for models it does not track. Counts fold the reasoning levels of one model together. Ranks compare each measure with every measured model in the catalogue, configurations included, as model pages do. The price is the published blended price (3 input tokens to 1 output token).
Figures computed by BenchSift from the catalogue: Artificial Analysis, OpenRouter, Models.dev and Hugging Face. A value no model carries is not shown.
Data retrieved on Nov 12, 2026.