Step 3.7 Flash
StepFun’s flagship multimodal reasoning model.
text + image → text · made by StepFun
Sold by 11 ways
| Seller | Lane | Rate |
|---|---|---|
| StepFun api | standard | $0.2per Mtok in$1.2per Mtok out$0.04per Mtok cachedplatform.stepfun.com · read 2026-08-25 |
| Nous Research aggregator | standard | $0.16per Mtok in$0.92per Mtok out$0.032per Mtok cachedinference-api.nousresearch.com · read 2026-09-11 |
| OpenRouter 2 ways | $0.2per Mtok in$1.15per Mtok out$0.04per Mtok cached | |
| OpenRouter aggregator | deepinfra | $0.16per Mtok in$0.92per Mtok out$0.032per Mtok cachedopenrouter.ai · read 2026-09-03 |
| OpenRouter aggregator | standard | $0.2per Mtok in$1.15per Mtok out$0.04per Mtok cachedopenrouter.ai · read 2026-08-24 |
| DeepInfra aggregator | standard | $0.2per Mtok in$1.15per Mtok outapi.deepinfra.com · read 2026-08-25 |
| ElectronHub aggregator | standard | $0.2per Mtok in$1.15per Mtok outapi.electronhub.ai · read 2026-09-16 |
| Hugging Face Inference Providers 2 ways | $0.2per Mtok in$1.15per Mtok out | |
| Hugging Face Inference Providers aggregator | standard | $0.2per Mtok in$1.15per Mtok outmodels.dev · read 2026-09-16 |
| Hugging Face Inference Providers aggregator | deepinfra | $0.2per Mtok in$1.15per Mtok outrouter.huggingface.co · read 2026-08-25 |
| Novita AI aggregator | standard | $0.2per Mtok in$1.15per Mtok out$0.04per Mtok cachedapi.novita.ai · read 2026-08-25 |
| Requesty aggregator | novita | $0.2per Mtok in$1.15per Mtok out$0.04per Mtok cachedrouter.requesty.ai · read 2026-08-28 |
| Vercel AI Gateway aggregator | standard | $0.2per Mtok in$1.15per Mtok out$0.04per Mtok cachedai-gateway.vercel.sh · read 2026-08-25 |
Measured 14 standings
| Place | Board | Metric | Score |
|---|---|---|---|
| 18thof 32 | AIME 2026 | Accuracy (± 95% CI) | 95 |
| 58thof 92 | Design Arena · asciiart (models) | Elo | 1164 |
| 72ndof 112 | Ale bench — Epoch AI | Performance | 694.12 |
| 73rdof 162 | Design Arena · website (models) | Elo | 1207 |
| 76thof 150 | Design Arena · uicomponent (models) | Elo | 1196 |
| 78thof 156 | Design Arena · codecategories (models) | Elo | 1196 |
| 87thof 147 | Design Arena · 3d (models) | Elo | 1156 |
| 87thof 152 | Design Arena · dataviz (models) | Elo | 1186 |
| 87thof 155 | Design Arena · gamedev (models) | Elo | 1177 |
| 88thof 109 | Design Arena · svg (models) | Elo | 1098 |
| 100thof 177 | Critpt — Epoch AI | Accuracy | 2.286 |
| 114thof 184 | Artificial Analysis · Coding Index | Index | 39.6 |
| 115thof 168 | Scicode — Epoch AI | Score | 40.046 |
| 123rdof 302 | Artificial Analysis · Intelligence Index | Artificial Analysis Intelligence Index | 19 |
About
Step 3.7 Flash — a text model from StepFun, sold by 9 companies from $0.16 in and $0.92 out per million tokens, placed 123rd of 302 on Artificial Analysis · Intelligence Index.
It takes text and images and returns text, with a context window of 256,000 tokens. It was published in May 2026, trained on material up to March 2026. Its sellers say it can reason step by step and call a tool. The catalogue files it under extract and reasoning. Nine companies sell it. The cheapest is $0.16 in and $0.92 out per million tokens at Nous Research. Beside the standard rate there is a separately routed lane. It has been measured on 14 boards, and stands best at 123rd of 302 on Artificial Analysis · Intelligence Index.
Every current figure
- Maker
- StepFun
- Register
- model
- Takes
- text + image
- Returns
- text
- Context
- 256,000 tokens
- Published
- May 2026
- Knowledge to
- March 2026
- Parameters
- 201 billion
- Licence
- not read
- Sellers
- 9
- Price
- $0.16 in and $0.92 out per million tokens — Nous Research
- Boards
- 14
- Best place
- 123rd of 302 — Artificial Analysis · Intelligence Index
Known as 4 names
stepfun-ai/Step-3.7-Flashstepfun/step-3.7-flashnovita/stepfun/step-3.7-flashstep-3.7-flash