StepFun: Step3
OpenAI • text • vision • function-calling • json-mode
Provider ID
stepfun-ai/step3Step3 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.
Quick Summary
Best For:
High-volume, low-latency tasks where cost efficiency is paramount
Pricing:
$0.00/1M input tokens, $0.00/1M output tokens
Context Window:
65,536 tokens
Key Differentiator:
Cost-optimized for high-volume usage
Specifications
Context Window
65,536 tokens
Max Output Tokens
65,536 tokens
Streaming
Yes
JSON Mode
Yes
Vision
Yes
Tier
Affordable
Capabilities
text
vision
function-calling
json-mode
Tags