Prime Intellect Releases INTELLECT-3: 106B-Parameter MoE Model Trained with Large-Scale RL and Fully Open-Source
Original title:INTELLECT-3: A 100B+ MoE trained with large-scale RL
AI summary
Prime Intellect released INTELLECT-3, a 106B-parameter MoE model trained on the GLM-4.5-Air base with SFT and large-scale RL, achieving state-of-the-art results at its scale on math, code, science and reasoning benchmarks.
Key Takeaways
- Prime Intellect released INTELLECT-3,a 106B-parameter MoE model trained on the GLM-4.
- 5-Air base with SFT and large-scale RL,achieving state-of-the-art results at its scale on math
- science and reasoning benchmarks.
Quick Facts
- Published
- Updated
- Source
- Prime Intellect(网页)
- Category
- Model Releases
- AI score
- 69
原文给出了完整的模型权重、训练框架与 RL 环境开源清单,读者可以据此复现大规模强化学习后训练。
Related entities
Source & copyright
This article is based on curated titles and AI summaries from the AIHOT public API, with attribution and links back. Original copyright remains with the original publishers and authors; this site does not mirror full texts.
How to Cite
When citing this article, keep the title, date, and permalink, and credit Orange News as the source.
Orange News: Prime Intellect Releases INTELLECT-3: 106B-Parameter MoE Model Trained with Large-Scale RL and Fully Open-Source. November 26, 2025. https://new.nat6.net/en/news/cmtym43k9000lro10m0azgkme/