llmboard.ai
Benchmarks
CompareRankings
llmboard.ai
Benchmarks
CompareRankings
HomeVendorsStepFun

Model vendor

StepFun

StepFun (阶跃星辰) is a Chinese artificial intelligence company focused on developing large language models and multimodal AI systems.

Updated Aug 10, 2026. Data as of 2026-08-07

Website
Published models2
Published versions4
Open defaults1
CountryCN

On this page

  • Models
  • Scores
  • Highlights
  • About
  • FAQ

StepFun models

Default-version specifications and scored-version capability values are labeled separately.

2 rows
Columns

Show columns

STStep 3.5 FlashStepFunStep-3.5-FlashN/AN/A256KOpen weightsFeb 2, 2026Not scored
STStep3 VL 10BStepFunStep3-VL-10BN/AN/AN/AClosed weightsJan 15, 2026Not scored

StepFun capability scores

Each unique model uses its latest scored version.

No vendor score

None of this vendor's published models currently has a calculated LLMBoard score.

StepFun highlights

Leaders from this vendor's current published catalog.

Largest contextStep 3.5 Flash256K tokens
Open-weight defaults1of 2 published models

What is StepFun?

Vendor context from the published model registry.

StepFun (阶跃星辰) is a Chinese artificial intelligence company focused on developing large language models and multimodal AI systems.

Model ownership is separate from API hosting. One model can be offered by multiple API providers with different prices.

Vendor
StepFun
Published models
2
Published versions
4
Country
CN
Founded
Not published

Vendor description and inventory come from the published model registry.

FAQ

Common questions about StepFun.

What is StepFun's highest-ranked model?

No model from this vendor currently has a published score.

How many StepFun models are tracked?

The current public catalog contains 2 independently selectable StepFun model products and 4 published versions.

Does StepFun host every listed model API?

Not necessarily. Vendor identifies the model developer; API providers are tracked separately.

Does StepFun publish open-weight models?

1 current default versions are marked as open weight.

Rankings

OverallCodingText ArenaPricing

Modalities

Image GenerationVideo GenerationSpeech-to-TextEmbeddings

Benchmarks

All BenchmarksReasoningMathCoding

Vendors

All VendorsOpenAIAnthropicGoogle
llmboard.aiCopyright 2026 llmboard.ai