Models/StepFun/Step 3.5 Flash

Step 3.5 Flash

StepFun 🧠 Language Model

StepFun's latest reasoning model — 196B MoE (11B active), 256K context, open-source Apache 2.0. Fast agentic intelligence with strong code and math.

Specifications

Context Window256,000 tokens
Max Output16,384 tokens
Speed●●●●● Fast
ModalitiesInput: text  ·  Output: text
FeaturesTools: Yes Streaming: Yes

Pricing

Included with plan

Capabilities

toolsreasoning
Use Step 3.5 Flash on Zubnet →

More from StepFun

S
Step 3
StepFun's frontier reasoning model — top-tier performance on math, code, and complex reasoning tasks. 66K context.
🧠 Language Model
S
Step 2 Mini
StepFun's compact text model — fast and cost-efficient with strong code ability. 32K context.
🧠 Language Model
S
Step 1o Turbo Vision
StepFun's vision-language reasoning model — accepts text and images with extended thinking for complex visual analysis.
🧠 Language Model
S
Step R1-V Mini
StepFun's compact vision reasoning model — multimodal input with reasoning capabilities at lower cost.
🧠 Language Model