Step 3.7 Flash
Timed at 100 tok/s, scored 39.6 on the Coding Index, and on 9 Design Arena boards.
retrieved sha256 · 0bb53a77713f
The same weights, served by one provider at its own list price, read from novita.ai. OpenRouter's rate is its routed rate across whichever sellers it uses, and the two need not agree.
Measured speed
2,340 requests · 2 endpoints · 30-minute windowOpenRouter's own timing of its routed traffic, weighted by requests, over one 30-minute window read . Every route is on the endpoint ledger below.
Every endpoint
2 priced · 2 timed · cheapest first- Novitanovita/fp8 · listed$0.20$1.15$0.04fp8256k35.5 tok/s15.33.92s75 req100%
- StepFunstepfun/fp8 · listed$0.20$1.15$0.04fp8230k103 tok/s273.72s2,265 req100%
Each track runs from a route's p90, the rate nine in ten of its requests met or beat, up to its median. Weighted by requests, the figure above reports where the traffic went rather than what one route does: 1 of the 2 timed routes are slower than it, and the middle route runs at 69 tok/s. The widest median-to-p90 gap is StepFun's, 3.8x.
Uptime is OpenRouter's own figure for each seller over the trailing day, read once a day, and the column prints the latest reading, taken . 2 of the 2 sellers with a reading fell under 99% on at least one of the 23 days held, 23 Oct 2026 to 21 Nov 2026: Novita at 97.94% on 18 Nov 2026 and StepFun at 98.58% on 10 Nov 2026. Source: OpenRouter's public endpoints API, one reading a day for each seller: 1,211 sellers carried one on the latest of the 23 days held.
Source: OpenRouter's public endpoints API. 1257 sellers across 359 models when the mirror ran, 186 of them served by more than one. List prices, not what anybody paid; a discount is the seller's own. Read . Speed is OpenRouter's own timing over one 30-minute window, read .
What each seller accepts
18 parameters · 2 sellers- frequency_penaltyevery seller
- include_reasoningevery seller
- logprobsevery seller
- max_tokensevery seller
- reasoningevery seller
- reasoning_effortevery seller
- stopevery seller
- structured_outputsevery seller
- temperatureevery seller
- tool_choiceevery seller
- toolsevery seller
- top_logprobsevery seller
- top_pevery seller
- presence_penalty1 of 2not StepFun
- repetition_penalty1 of 2not StepFun
- response_format1 of 2not Novita
- seed1 of 2not StepFun
- top_k1 of 2not StepFun
13 of these 18 are accepted by every one of the 2 sellers that publish a list; the other 5 are declared by some sellers and not others, so what a request sending one of them gets depends on which seller serves it. Each list is the seller's own declaration in OpenRouter's endpoints API, read , and nothing here tested one.
Price position
342 priced modelsBlended $0.438 per million tokens, 3 in to 1 out: cheaper than 225 of the 342 priced models in the catalog. Each bar is one stretch of the catalog on a logarithmic price scale; the mark is this listing.
Measured quality
145 models with an index · 132 on a design arena boardEach mark is one board this model stands on, placed at the share of that board it scores above, and led to its number in the list under the rule. The columns and the ledger below carry the figures.
- pdf · Agents1152 Elowin rate 44.1% · source rank 27the board, 36 ranked
- 3D scenes · Models1108 Elowin rate 39.6% · source rank 91the board, 113 ranked
- ASCII art · Models1132 Elowin rate 44.3% · source rank 49the board, 67 ranked
- Code · Models1168 Elowin rate 41.7% · source rank 74the board, 120 ranked
- Data visualisation · Models1172 Elowin rate 43.3% · source rank 69the board, 110 ranked
- Game development · Models1143 Elowin rate 37.1% · source rank 79the board, 120 ranked
- SVG · Models1072 Elowin rate 36.2% · source rank 75the board, 83 ranked
- UI components · Models1165 Elowin rate 41.3% · source rank 75the board, 116 ranked
- Websites · Models1185 Elowin rate 43.7% · source rank 75the board, 126 ranked
Epoch AI has not placed Step 3.7 Flash on its Capabilities Index, or places it under a name the pairing does not reach. Epoch places a model once four of its benchmarks carry a result.
Benchmark results
none of 6 boardsNone of the 6 openly licensed boards Vitna mirrors has published a result for Step 3.7 Flash under a name that matches its listing, and names are matched to the catalog exactly and never guessed. Every board, with its licence and what its publisher measures, is on /benchmarks.
Source: OpenRouter (openrouter.ai/models), as of 17 Nov 16:29 UTC. sha256 0bb53a77713f over 853,270 bytes.
Catalog: OpenRouter's public model list.
Intelligence, coding and agentic indexes: Artificial Analysis, as published by OpenRouter.
Elo, win rate and rank: Design Arena, as published by OpenRouter.
Source: each model's own Hugging Face repository (huggingface.co), as of 20 Nov 2026. 162 repositories read, 4 did not answer.
Licence, parameter count, model class and gating: each model's own Hugging Face repository, the one OpenRouter names for the listing.
A licence is the tag the uploader put on the repository, read from nothing else and verified by nobody.
A parameter count is the total the repository's safetensors files declare; for a mixture of experts that is every expert, not the share one token runs through.
Vitna ran none of these evaluations and adjusted none of these figures.
