Step 3.5 Flash
StepFun · released Feb 1, 2026 · stepfun-ai/Step-3.5-Flash
- Type
- Open weightsApache License 2.0
- Params
- 199B
- Context
- 262K
active per word not recorded by us · about 197K words of context
Our take
Written Sep 3, 2026Step 3.5 Flash is a 199.4-billion-parameter text model from StepFun with a permissive Apache licence and a 262,144-token request limit. Its coding skill scores highest among its measured dimensions, though creative writing lags, and only two hosts currently offer it at identical rates.
Pick this for Apache-licensed self-hosting or research at large scale, or for coding generation where human preference rankings guide your choice. Use it if you want broad arena evaluation across six measured skills rather than a single headline score. Skip it if you need image or audio input, if creative writing quality is critical, or if you want active-parameter efficiency data to verify inference cost claims.
The case for it
- Coding is its standout measured skill, scoring 56.1 points above its own overall text score.
- Apache 2.0 licence allows commercial use, modification and redistribution without restriction.
- Six Arena benchmarks cover coding, creative writing, hard prompts, instruction following, maths and overall text.
The case against it
- No disclosed active-parameter count, so efficiency claims about its 199.4 billion total cannot be verified.
- Creative writing scores 49.1 points below its own overall text score, the weakest of its measured skills.
- Only two tracked offers with identical pricing, so there is no price competition.
How good is it?
EverydayGeneral questions and everyday reasoning
Arena Text (overall)102nd of 168 · 1393
CodingWriting and fixing code on its own
Arena Coding94th of 168 · 1449
Arena Coding is the only board that has scored it for this.
AgenticPlanning, calling tools, staying on task
Not yet scored on Arena Agent.
WritingDrafting and rewriting prose
Arena Creative Writing104th of 168 · 1345
Arena Creative Writing is the only board that has scored it for this.
These tests check whether a model follows instructions — a precondition for all the work above, but not a measure of how well that work is done.
Every published score for this model6 scoresEvery figure we hold, from 6 boards, with who ran it and a link to the source — including the boards no rating above is built on.
Can you run it yourself?
GeForce RTX 4090 · 24 GB
Too large for this card. The weights do not fit even with part of them offloaded to system memory.
Apple M1 Pro (16-core GPU) · 32 GB
Too large for this card. The weights do not fit even with part of them offloaded to system memory.
Apple M2 Ultra (76-core GPU) · 192 GB
Borderline fit on an estimated size. It leaves 12.8 GB spare on a size we calculated rather than measured, and a 10% error either way would change the answer.
Memory use by level
Against a 24 GB card.
What is quantisation? →This model on every device we track71 devicesThe Q4 build most people download, on each device: what the weights come to, how much context the memory leaves, and whether it runs. Smallest device that runs it first.
Check against your own machine → · Where to rent it hosted →
Or rent it from someone else
Prices checked 4 hours ago — each listing carries its own date.
- per 1M tokens
- $0.10 in / $0.30 out
- Context served
- 262K
- Throughput
- Not measured
| Provider | In / out per 1M tokens | Context | Throughput | Trains on prompts | Logs prompts | Zero retention |
|---|---|---|---|---|---|---|
| OpenRouterOpenRouter's own listing | $0.10 / $0.30checked 4 hours ago | 262K | not measured | Unknown | Unknown | Unknown |
| SiliconFlowfp8Through OpenRouter | $0.10 / $0.30checked 4 hours ago | 262K66K max reply | 69 tok/s | No | No | Confirmed |
Across the 2 listings we hold: 1 says it does not train on prompts, 0 say they do and 1 does not say. 1 appears in the zero-retention registry we check; the rest are unknown to us.
What each host's API supports
From the parameter list each endpoint publishes. Streaming is omitted: nothing we hold reports it, for any model.
| Provider | Tool calling | JSON output | Strict schema |
|---|---|---|---|
| OpenRouterOpenRouter's own listing | ✓ | ✗ | ✗ |
| SiliconFlowfp8Through OpenRouter | ✓ | ✗ | ✗ |
Tool calling: 2 of 2 listings say yes. JSON output: 0 of 2 listings say yes, 2 say no. Strict schema: 0 of 2 listings say yes, 2 say no.
Models people weigh against Step 3.5 Flash
When we formed this view
Recent changes
What moved
first indexed by our pipelineEach date is the day we first saw the change, or the day the maker announced it.
What we do not know about this model yet
- We hold no measured file for it, so all 3 sizes on this page are calculated from the parameter count.
- We do not hold the active parameter count for it, so how much of it runs on any one token is unknown to us.
- Nothing we hold says whether an endpoint streams, so we do not show it either way.
- 1 of 2 listings does not say whether it trains on prompts.
- We hold no cached-input rate for any of its listings.
- We hold no batch or off-peak rate for any of its listings.
- We hold a decode speed for it, but no prompt-processing (prefill) figure, so how long the input side of a job takes is unknown to us.
Licence and identifiers
What the licence allowsApache License 2.0, what it allows commercially, and the identifiers you need to pull this model — its Hugging Face repo, our slug and a machine-readable card.
Licence
Apache License 2.0
Fully permissive: commercial use, redistribution, and derivatives allowed. Requires attribution and a copy of the license. Includes an express patent grant.
Identifiers
- Hugging Face
- stepfun-ai/Step-3.5-Flash
- Architecture
- Mixture of experts
- Takes in, gives back
- Text in, text out
- Catalogue slug
- stepfun-step-3-5-flash