Step 3.5 Flash Benchmark Scores & Performance

Benchmark analysis of Step 3.5 Flash by StepFun across 32 sourced tests on BenchLM.

According to BenchLM.ai, Step 3.5 Flash ranks #40 out of 123 models with an overall score of 66/100. While not a frontier model, it offers specific advantages depending on the use case.

Step 3.5 Flash is a open weight model with a 256K token context window. It processes queries without explicit chain-of-thought reasoning, offering faster response times and lower token usage.

Its strongest category is Multilingual (#28), while its weakest is Multimodal & Grounded (#64). This performance profile makes it a well-rounded choice across a range of tasks.

Creator

StepFun

Source Type

Open Weight

Reasoning

Non-Reasoning

Context Window

256K

Overall Score

66#40 of 123

Arena Elo

1266

Knowledge Benchmarks

MMLU
84
GPQA
83
SuperGPQA
81
OpenBookQA
79
MMLU-Pro
76
HLE
12
FrontierScience
71

Coding Benchmarks

HumanEval
77
SWE-bench Verified
48
LiveCodeBench
45
SWE-bench Pro
49

Mathematics Benchmarks

AIME 2023
85
AIME 2024
87
AIME 2025
86
HMMT Feb 2023
81
HMMT Feb 2024
83
HMMT Feb 2025
82
BRUMO 2025
84
MATH-500
85

Reasoning Benchmarks

SimpleQA
82
MuSR
82
BBH
83
LongBench v2
74
MRCRv2
73

Agentic Benchmarks

Terminal-Bench 2.0
62
BrowseComp
66
OSWorld-Verified
54

Multimodal & Grounded Benchmarks

MMMU-Pro
64
OfficeQA Pro
70

Instruction Following Benchmarks

IFEval
87

Multilingual Benchmarks

MGSM
86
MMLU-ProX
81

Frequently Asked Questions

How does Step 3.5 Flash perform overall in AI benchmarks?

Step 3.5 Flash ranks #40 out of 123 models with an overall score of 66. It is created by StepFun and features a 256K context window.

Is Step 3.5 Flash good for knowledge and understanding?

Step 3.5 Flash ranks #43 out of 123 models in knowledge and understanding benchmarks with an average score of 60.8. There are stronger options in this category.

Is Step 3.5 Flash good for coding and programming?

Step 3.5 Flash ranks #39 out of 123 models in coding and programming benchmarks with an average score of 47.1. There are stronger options in this category.

Is Step 3.5 Flash good for mathematics?

Step 3.5 Flash ranks #37 out of 123 models in mathematics benchmarks with an average score of 84.5. There are stronger options in this category.

Is Step 3.5 Flash good for reasoning and logic?

Step 3.5 Flash ranks #39 out of 123 models in reasoning and logic benchmarks with an average score of 78.3. There are stronger options in this category.

Is Step 3.5 Flash good for agentic tool use and computer tasks?

Step 3.5 Flash ranks #45 out of 123 models in agentic tool use and computer tasks benchmarks with an average score of 60.2. There are stronger options in this category.

Is Step 3.5 Flash good for multimodal and grounded tasks?

Step 3.5 Flash ranks #64 out of 123 models in multimodal and grounded tasks benchmarks with an average score of 66.7. There are stronger options in this category.

Is Step 3.5 Flash good for instruction following?

Step 3.5 Flash ranks #33 out of 123 models in instruction following benchmarks with an average score of 87. There are stronger options in this category.

Is Step 3.5 Flash good for multilingual tasks?

Step 3.5 Flash ranks #28 out of 123 models in multilingual tasks benchmarks with an average score of 82.8. There are stronger options in this category.

Is Step 3.5 Flash open source?

Yes, Step 3.5 Flash is an open weight model created by StepFun, meaning it can be downloaded and run locally or fine-tuned for specific use cases.

What is the context window size of Step 3.5 Flash?

Step 3.5 Flash has a context window of 256K, which determines how much text it can process in a single interaction.

Last updated: March 12, 2026

Weekly LLM Updates

New model releases, benchmark scores, and leaderboard changes. Every Friday.

Free. Your signup is stored with a derived country code for compliance routing.