StepFun Launches Step 5 Preview: A 600B-Total, 27B-Active MoE Model With 1M Context for Long-Horizon Agentic Work
StepFun launched Step 5 Preview, a 600B MoE model with 1M context for agentic work, claiming cost efficiency with open weights coming in October.
StepFun has released Step 5 Preview, its new flagship Mixture-of-Experts model for agentic work, featuring 600B total parameters and a 1M-token context window. The model activates approximately 27B parameters per token, offering a cost-effective alternative for software engineering and knowledge work. StepFun claims competitive benchmarks, though independent analysis by Artificial Analysis scores it below frontier models like Claude Opus 5 and GPT-6 Astra on intelligence.
- StepFun launched Step 5 Preview, a 600B-total, 27B-active MoE model.
- The model offers a 1M-token context window and text/image/video input.
- API pricing is $1.00 per 1M input tokens and $2.70 per 1M output tokens.
- Open weights are scheduled for release on October 15, 2026.
- Independent benchmarks score it 44 on Artificial Analysis's Intelligence Index.
- Organizations
- StepFun
- AI models
- Step 5 Preview
From the original article
3 min read at marktechpost.com
StepFun has released Step 5 Preview, a sparse Mixture-of-Experts model with 600B total parameters and 27B active per token. It supports a 1M-token context window and accepts text, image, and video input. The model targets long-horizon agentic work in software engineering, professional knowledge work, and finance.
By Michal Sutter · marktechpost.com
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.