StepFunLanguage model
Step 3.5 Flash 2603
Step 3.5 Flash 2603 has no overall index value yet because there are too few independent benchmark results.
Intelligencenot ranked yet
n/aOverall indexhigher is better
SpeedRank 36 of 101
107Tokens per secondhigher is better
Priceno API price
n/aper 1M tokenslower is better
ContextRank 77 of 103
256Ktokens at once≈ 380 A4 pages
PopularityRank 710 of 785
19out of 100+1 vs. last month
based on two data groups, mostly search
Summary
At 107 tokens per second it writes faster than a typical model (87).
Step 3.5 Flash 2603 is StepFun’s large sparse Mixture-of-Experts foundation model for reasoning, coding, and agentic work. Its official card lists 196.81 billion total parameters and about 11 billion active per token. MTP-3 and a 256,000-token context are intended to combine fast generation with long workflows. This checkpoint is for text-to-text open Apache-2.0 weights, not a multimodal Step configuration.
StepFun currently recommends a nightly vLLM path and notes that full MTP-3 support is not yet available there. SGLang, Transformers, and llama.cpp are also documented. API access through StepFun and OpenRouter is listed, while local serving has to account for the 196.81 billion total parameters and the backend support state.
All benchmark results
Every published result with its source, test setup and comparison models.
Show all 25 published benchmark results
Page 1 of 3
Measurements without matching peer values
These measurements have no matching peer values under the same test conditions. Their original values and sources remain available here.
CritPt v2: Overall (Show measurement, test conditions, and source)
- Source value
- 2.28571428571429
- Score
- 2.28571428571429
- Metric
- pass@1 accuracy
- Unit
- %
- Benchmark version
- v2
- Category
- reasoning
- Task
- overall
- Task label
- Overall
- Direction
- Higher is better
- Source type
- independent-evaluator
- Evaluator
- Artificial Analysis
- Status
- disputed
- Retrieved at
- 2026-11-02
- Methodology
- Exact AA configuration: Step 3.5 Flash 2603 (release step-3-5-flash, row step-3-5-flash, UUID 1aa3694e-b656-4dbe-8f84-0c65d8897abb, reasoning=true, effort=not specified, suffix=not specified). AA source value for average reasoning tokens before the answer: 2000. Its methodology averages 60 varied prompts and assumes 2,000 when no average is available. This is not a token cap. The full CritPt v2 run covers 70 questions with five repeats. The official grading server reports pass@1.. Source model ID: artificial-analysis/1aa3694e-b656-4dbe-8f84-0c65d8897abb/step-3-5-flash/step-3-5-flash. Source model name: Step 3.5 Flash 2603. Provider reported by source: StepFun. Model snapshot: Step 3.5 Flash 2603. Evaluator: Artificial Analysis. Dataset type: artificial-analysis-evaluation. Tool policy: Python functions, symbolic expressions, and numerical answers; official grading server. Attempts: 70 questions, five repeats per question, pass@1. Reasoning: reasoning enabled. Score display: 2.3%. Source ID: artificial-analysis. Evidence capture ID: aa-model-page-2026-11-02-67d46be20a5b. Source payload hash: 67d46be20a5b412db6cfb60ad190441fcaa5513500b8fd47d033168d0e97b6b2. Parser version: artificial-analysis-rsc-models-v3. Imported at: 2026-11-02. Source cohort: artificial-analysis-critpt-v2:v2:standalone:reasoning:effort=unspecified:suffix=unspecified
- Comparison methodology
- Exact AA configuration: Step 3.5 Flash 2603 (release step-3-5-flash, row step-3-5-flash, UUID 1aa3694e-b656-4dbe-8f84-0c65d8897abb, reasoning=true, effort=not specified, suffix=not specified). AA source value for average reasoning tokens before the answer: 2000. Its methodology averages 60 varied prompts and assumes 2,000 when no average is available. This is not a token cap. The full CritPt v2 run covers 70 questions with five repeats. The official grading server reports pass@1.. Tool policy: Python functions, symbolic expressions, and numerical answers; official grading server. Attempts: 70 questions, five repeats per question, pass@1. Reasoning: reasoning enabled
- Context
- This AA source field contains the original CritPt score. The earlier CritPt v2 attribution was incorrect; the value and former label remain as disputed historical evidence.. The source dataset marks this result as disputed. It remains visible as documented evidence, but is not reliable comparative proof.
IFBench: Overall (Show measurement, test conditions, and source)
- Source value
- 66.53061224489801
- Score
- 66.53061224489801
- Metric
- IFBench prompt-level loose accuracy (pass@1)
- Unit
- %
- Benchmark version
- AA test set: 294 questions
- Category
- general
- Task
- overall
- Task label
- Overall
- Direction
- Higher is better
- Source type
- independent-evaluator
- Evaluator
- Artificial Analysis
- Status
- superseded
- Retrieved at
- 2026-11-02
- Methodology
- Exact AA configuration: Step 3.5 Flash 2603 (release step-3-5-flash, row step-3-5-flash, UUID 1aa3694e-b656-4dbe-8f84-0c65d8897abb, reasoning=true, effort=not specified, suffix=not specified). AA source value for average reasoning tokens before the answer: 2000. Its methodology averages 60 varied prompts and assumes 2,000 when no average is available. This is not a token cap. Artificial Analysis runs 294 questions from the IFBench test set with five repeats per question. The official evaluator scores prompt-level accuracy in loose mode as pass@1.. Source model ID: artificial-analysis/1aa3694e-b656-4dbe-8f84-0c65d8897abb/step-3-5-flash/step-3-5-flash. Source model name: Step 3.5 Flash 2603. Provider reported by source: StepFun. Model snapshot: Step 3.5 Flash 2603. Evaluator: Artificial Analysis. Dataset type: artificial-analysis-evaluation. Tool policy: Single-turn answer evaluated with the official IFBench source code. Attempts: pass@1, 5 repeats per question. Reasoning: reasoning enabled. Score display: 66.5%. Source ID: artificial-analysis. Evidence capture ID: aa-model-page-2026-11-02-67d46be20a5b. Source payload hash: 67d46be20a5b412db6cfb60ad190441fcaa5513500b8fd47d033168d0e97b6b2. Parser version: artificial-analysis-rsc-models-v3. Imported at: 2026-11-02. Source cohort: artificial-analysis-ifbench:AA test set: 294 questions:standalone:reasoning:effort=unspecified:suffix=unspecified
- Comparison methodology
- Exact AA configuration: Step 3.5 Flash 2603 (release step-3-5-flash, row step-3-5-flash, UUID 1aa3694e-b656-4dbe-8f84-0c65d8897abb, reasoning=true, effort=not specified, suffix=not specified). AA source value for average reasoning tokens before the answer: 2000. Its methodology averages 60 varied prompts and assumes 2,000 when no average is available. This is not a token cap. Artificial Analysis runs 294 questions from the IFBench test set with five repeats per question. The official evaluator scores prompt-level accuracy in loose mode as pass@1.. Tool policy: Single-turn answer evaluated with the official IFBench source code. Attempts: pass@1, 5 repeats per question. Reasoning: reasoning enabled
- Context
- The often cited 58 refers to verifiable out-of-domain constraint types, not the number of questions in AA's test run.. This result is retained as historical evidence and is not interchangeable with current results.
IFBench: Overall (Show measurement, test conditions, and source)
- Source value
- 66.53061224489801
- Score
- 66.53061224489801
- Metric
- IFBench prompt-level loose accuracy (pass@1)
- Unit
- %
- Benchmark version
- AA test set: 294 questions
- Category
- general
- Task
- overall
- Task label
- Overall
- Direction
- Higher is better
- Source type
- independent-evaluator
- Evaluator
- Artificial Analysis
- Status
- active
- Retrieved at
- 2026-11-07
- Methodology
- Exact AA configuration: Step 3.5 Flash 2603 (release step-3-5-flash, row step-3-5-flash, UUID 1aa3694e-b656-4dbe-8f84-0c65d8897abb, reasoning=true, effort=not specified, suffix=not specified). This source capture does not provide an average reasoning-token count. Artificial Analysis runs 294 questions from the IFBench test set with five repeats per question. The official evaluator scores prompt-level accuracy in loose mode as pass@1.. Source model ID: artificial-analysis/1aa3694e-b656-4dbe-8f84-0c65d8897abb/step-3-5-flash/step-3-5-flash. Source model name: Step 3.5 Flash 2603. Provider reported by source: StepFun. Model snapshot: Step 3.5 Flash 2603. Evaluator: Artificial Analysis. Dataset type: artificial-analysis-evaluation. Tool policy: Single-turn answer evaluated with the official IFBench source code. Attempts: pass@1, 5 repeats per question. Reasoning: reasoning enabled. Score display: 66.5%. Source ID: artificial-analysis. Evidence capture ID: aa-model-page-2026-11-07-b96e3ffc8c5b. Source payload hash: b96e3ffc8c5b052aa56f96d05e89d02b8cc234a75085820fca0ca2a39e498274. Parser version: artificial-analysis-rsc-models-v4. Imported at: 2026-11-07. Source cohort: artificial-analysis-ifbench:AA test set: 294 questions:standalone:reasoning:effort=unspecified:suffix=unspecified
- Comparison methodology
- Exact AA configuration: Step 3.5 Flash 2603 (release step-3-5-flash, row step-3-5-flash, UUID 1aa3694e-b656-4dbe-8f84-0c65d8897abb, reasoning=true, effort=not specified, suffix=not specified). This source capture does not provide an average reasoning-token count. Artificial Analysis runs 294 questions from the IFBench test set with five repeats per question. The official evaluator scores prompt-level accuracy in loose mode as pass@1.. Tool policy: Single-turn answer evaluated with the official IFBench source code. Attempts: pass@1, 5 repeats per question. Reasoning: reasoning enabled
- Context
- The often cited 58 refers to verifiable out-of-domain constraint types, not the number of questions in AA's test run.
SWE-bench Verified (Show measurement, test conditions, and source)
- Source value
- 74.4
- Score
- 74.4
- Metric
- resolved rate
- Unit
- %
- Benchmark version
- SWE-bench Verified
- Category
- coding
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- StepFun Step-3.5-Flash official model card
- Status
- active
- Retrieved at
- 2026-10-15
- Methodology
- Vendor-reported result. Exact Step 3.5 Flash model row/widget in StepFun Step-3.5-Flash official model card. Benchmark prompt, sampling/decoding settings, reasoning effort, quantization, and task run date are not stated in the cited result; none are inferred.
Terminal-Bench (Show measurement, test conditions, and source)
- Source value
- 51
- Score
- 51
- Metric
- score
- Unit
- %
- Benchmark version
- Terminal-Bench 2.0
- Category
- coding
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- StepFun Step-3.5-Flash official model card
- Status
- active
- Retrieved at
- 2026-10-15
- Methodology
- Vendor-reported result. Exact Step 3.5 Flash model row/widget in StepFun Step-3.5-Flash official model card. Benchmark prompt, sampling/decoding settings, reasoning effort, quantization, and task run date are not stated in the cited result; none are inferred.
τ²-Bench (Show measurement, test conditions, and source)
- Source value
- 88.2
- Score
- 88.2
- Metric
- score
- Unit
- No unit provided
- Benchmark version
- τ²-Bench
- Category
- agent
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- StepFun Step-3.5-Flash official model card
- Status
- active
- Retrieved at
- 2026-10-15
- Methodology
- Vendor-reported result. Exact Step 3.5 Flash model row/widget in StepFun Step-3.5-Flash official model card. Benchmark prompt, sampling/decoding settings, reasoning effort, quantization, and task run date are not stated in the cited result; none are inferred.
- Context
- The source presents a raw score without an explicit unit or scale; the published value is retained without percentage conversion.
BrowseComp (Show measurement, test conditions, and source)
- Source value
- 51.6
- Score
- 51.6
- Metric
- score
- Unit
- No unit provided
- Benchmark version
- BrowseComp
- Category
- agent
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- StepFun Step-3.5-Flash official model card
- Status
- active
- Retrieved at
- 2026-10-15
- Methodology
- Vendor-reported result. Exact Step 3.5 Flash model row/widget in StepFun Step-3.5-Flash official model card. Benchmark prompt, sampling/decoding settings, reasoning effort, quantization, and task run date are not stated in the cited result; none are inferred.
- Context
- The source presents a raw score without an explicit unit or scale; the published value is retained without percentage conversion.
BrowseComp with Context Manager (Show measurement, test conditions, and source)
- Source value
- 69
- Score
- 69
- Metric
- score
- Unit
- No unit provided
- Benchmark version
- BrowseComp with Context Manager
- Category
- agent
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- StepFun Step-3.5-Flash official model card
- Status
- active
- Retrieved at
- 2026-10-15
- Methodology
- Vendor-reported result. Exact Step 3.5 Flash model row/widget in StepFun Step-3.5-Flash official model card. Benchmark prompt, sampling/decoding settings, reasoning effort, quantization, and task run date are not stated in the cited result; none are inferred.
- Context
- The source presents a raw score without an explicit unit or scale; the published value is retained without percentage conversion.
BrowseComp-ZH (Show measurement, test conditions, and source)
- Source value
- 66.9
- Score
- 66.9
- Metric
- score
- Unit
- No unit provided
- Benchmark version
- BrowseComp-ZH
- Category
- agent
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- StepFun Step-3.5-Flash official model card
- Status
- active
- Retrieved at
- 2026-10-15
- Methodology
- Vendor-reported result. Exact Step 3.5 Flash model row/widget in StepFun Step-3.5-Flash official model card. Benchmark prompt, sampling/decoding settings, reasoning effort, quantization, and task run date are not stated in the cited result; none are inferred.
- Context
- The source presents a raw score without an explicit unit or scale; the published value is retained without percentage conversion.
BrowseComp-ZH with Context Manager (Show measurement, test conditions, and source)
- Source value
- 73.7
- Score
- 73.7
- Metric
- score
- Unit
- No unit provided
- Benchmark version
- BrowseComp-ZH with Context Manager
- Category
- agent
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- StepFun Step-3.5-Flash official model card
- Status
- active
- Retrieved at
- 2026-10-15
- Methodology
- Vendor-reported result. Exact Step 3.5 Flash model row/widget in StepFun Step-3.5-Flash official model card. Benchmark prompt, sampling/decoding settings, reasoning effort, quantization, and task run date are not stated in the cited result; none are inferred.
- Context
- The source presents a raw score without an explicit unit or scale; the published value is retained without percentage conversion.
GAIA without file (Show measurement, test conditions, and source)
- Source value
- 84.5
- Score
- 84.5
- Metric
- score
- Unit
- No unit provided
- Benchmark version
- GAIA no file
- Category
- agent
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- StepFun Step-3.5-Flash official model card
- Status
- active
- Retrieved at
- 2026-10-15
- Methodology
- Vendor-reported result. Exact Step 3.5 Flash model row/widget in StepFun Step-3.5-Flash official model card. Benchmark prompt, sampling/decoding settings, reasoning effort, quantization, and task run date are not stated in the cited result; none are inferred.
- Context
- The source presents a raw score without an explicit unit or scale; the published value is retained without percentage conversion.
xbench-DeepSearch (Show measurement, test conditions, and source)
- Source value
- 83.7
- Score
- 83.7
- Metric
- score
- Unit
- No unit provided
- Benchmark version
- xbench-DeepSearch 2025.05
- Category
- agent
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- StepFun Step-3.5-Flash official model card
- Status
- active
- Retrieved at
- 2026-10-15
- Methodology
- Vendor-reported result. Exact Step 3.5 Flash model row/widget in StepFun Step-3.5-Flash official model card. Benchmark prompt, sampling/decoding settings, reasoning effort, quantization, and task run date are not stated in the cited result; none are inferred.
- Context
- The source presents a raw score without an explicit unit or scale; the published value is retained without percentage conversion.
Page 1 of 3
Popularity compared
How often the model is searched for, used and mentioned, compared with every model of its category.
Popularity index
The 100 most popular models of this category, 0 to 100, higher is more popular
20 of 101 models
As of November 2026 · Source: gradually.ai popularity index
Over time
Weekly values from October 18, 2026 to November 7, 2026
Step 3.5 Flash 2603Codestral-22B-v0.1 19DBRX 19Llama 3.1 70B 19Voxtral Small (latest) 19
Method changes in the timeline, a jump at these points is no market event:
- 1November 1, 2026: Value only with search or usage data, 30-day attention
What is gaining popularity right now, across models, vendors, harnesses and AI subscriptions, is collected in All AI trends.
Similar models
More models of the same kind as Step 3.5 Flash 2603, chosen by model line, developer, size, price and release date.
All specifications
Architecture
Input and output
Watermarking and provenance
AI watermarking refers to invisible signals embedded in the output. Visible marks and provenance metadata are listed separately. Unknown means there is insufficient evidence for that particular field.
- AI watermarking
- Unknown
- no source
- Watermarking technique
- Not known
- no source
Licence and access
Prices
Not included in any subscription in Germany.
Frequently asked questions about Step 3.5 Flash 2603
What is the knowledge cutoff of Step 3.5 Flash 2603?
January 2025, according to StepFun. That is 15 months before release.
Stated by StepFun
JanAprJulOctJanAprKnowledge cutoffJanuary 2025ReleasedApr 2, 2026
How large is the context window of Step 3.5 Flash 2603?
256,000 tokens. That is how much text Step 3.5 Flash 2603 reads at once, about 380 A4 pages.
Page count as a rule of thumb from Artificial Analysis
Kimi K2.6262K
Qwen 3.8 27B262K
Kimi K2.7 Code262K
Kimi K2.5262K
Nemotron 3 Ultra262K
Kimi K2 Thinking262K
Step 3.5 Flash 2603Rank 77 of 103256K
How fast is Step 3.5 Flash 2603?
107 tokens per second. That is faster than a typical model at 87.
Measured by Artificial Analysis on Nov 6, 2026. AA 3.0.0, including time to first token, averaged across available workloads and five days.
GPT-6 Sol111
o3110
o4-mini110
Step 3.5 Flash 2603Rank 36 of 101107
Claude Opus 5.5106
Claude Sonnet 5103
Grok 4.20 Reasoning102
Can I run Step 3.5 Flash 2603 locally?
Yes, the weights are open. With 196.8B parameters it needs a lot of memory, more a server than a laptop.
Stated by StepFun
- Open weights
- yes
- License
- Apache 2.0
- Parameters
- 196.8B
- active per token
- 11B
Sources and data date
Show all 19 sources
Benchmark sources at a glance
Which independent evaluators tested this model. This is coverage, not a quality ranking.
| Evaluator | Test series | Composite indices | Recorded active measurements |
|---|---|---|---|
| Artificial Analysis | 7 | 0 | 8 |
Without vendor claims and historical or disputed results; those stay labelled in the single values.
| Type | Evidence and data date |
|---|---|
| Speed | StepFun, Reasoning enabled; effort unspecified. AA 3.0.0, including time to first token, averaged across available workloads and five days. Retrieved November 6, 2026. |
| Input | Documented modalities. Retrieved October 27, 2026. |
| Output | Documented modalities. Retrieved October 27, 2026. |
| Research status | Research date October 15, 2026. 9 source URLs checked. This documents the inspected sources, not an exhaustive inventory of every publication. |
| Not published | benchmark-evaluation-setup Note The card gives task/version names and explicitly calls SWE-bench Verified and Terminal-Bench values percentages, but does not publish general decoding parameters, quantization, or most task run dates. |
| Additional source | Hugging Face (retrieved October 15, 2026; October 19, 2026) · Step 3.5 Flash model card · Editorial description reviewed October 2, 2026 |
| Additional source | Artificial Analysis (retrieved November 2, 2026) |
| Additional source | Artificial Analysis (retrieved October 17, 2026; November 2, 2026; November 7, 2026) |
| Additional source | Model metadata (retrieved October 17, 2026)123456 |
| Additional source | Step 3.5 Flash technical paper · Editorial description reviewed October 2, 2026 |
| Additional source | OpenRouter: Popularity index (derived), CC BY 4.0 |
| Additional source | Vercel AI Gateway: Popularity index (derived), CC BY 4.0 |
| Additional source | Ollama: cumulative downloads for the popularity index |
| Additional source | Replicate: cumulative runs for the popularity index |
| Additional source | Google Trending Now: limited feed selection for Germany and the US |
| Additional source | Creative Commons Attribution 4.0 |