LangMart: StepFun: Step3
Model Overview
| Property | Value |
|---|---|
| Model ID | openrouter/stepfun-ai/step3 |
| Name | StepFun: Step3 |
| Provider | stepfun-ai |
| Released | 2025-08-28 |
Description
Step3 is a cutting-edge multimodal reasoning model—built on a Mixture-of-Experts architecture with 321B total parameters and 38B active. It is designed end-to-end to minimize decoding costs while delivering top-tier performance in vision–language reasoning. Through the co-design of Multi-Matrix Factorization Attention (MFA) and Attention-FFN Disaggregation (AFD), Step3 maintains exceptional efficiency across both flagship and low-end accelerators.
Description
LangMart: StepFun: Step3 is a language model provided by stepfun-ai. This model offers advanced capabilities for natural language processing tasks.
Provider
stepfun-ai
Specifications
| Spec | Value |
|---|---|
| Context Window | 65,536 tokens |
| Modalities | text+image->text |
| Input Modalities | image, text |
| Output Modalities | text |
Pricing
| Type | Price |
|---|---|
| Input | $0.57 per 1M tokens |
| Output | $1.42 per 1M tokens |
Capabilities
- Frequency penalty
- Include reasoning
- Reasoning
- Response format
- Structured outputs
- Temperature
- Tool choice
- Tools
- Top k
- Top p
Detailed Analysis
StepFun: Step3