Ant Group releases trillion-parameter reasoning model Ring-2.6-1T
Original title:Thanks @AdinaYakup and the @huggingface community for the continued recognition! We feel happy to b...
AI summary
Ant Group has released Ring-2.6-1T, a trillion-parameter reasoning model built specifically for real-world agent workflows. The model is released under the MIT license, with context length extended from 128K to 256K via YaRN technology, and uses a hybrid training architecture combining Async RL and IcePop. Its core feature is dual reasoning effort modes: "high" mode for fast agent loops and "xhigh" mode for deep reasoning, aiming for a better balance between cost and performance. The model is open-source, and community feedback is welcome.
Key Takeaways
- Ant Group has released Ring-2.
- 6-1T,a trillion-parameter reasoning model built specifically for real-world agent workflows.
- The model is released under the MIT license,with context length extended from 128K to 256K via…
- Its core feature is dual reasoning effort modes: "high" mode for fast agent loops and "xhigh" mode…
- The model is open-source,and community feedback is welcome.
Quick Facts
- Published
- Updated
- Source
- X:蚂蚁百灵 (@AntLingAGI)
- Category
- Model Releases
- AI score
- 76
蚂蚁把1T推理模型以MIT许可开源,还专门为Agent循环做了高低推理双模式,做开源Agent的团队直接多了一个强力基座可选。
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: Ant Group releases trillion-parameter reasoning model Ring-2.6-1T. May 17, 2026. https://new.nat6.net/en/news/cmp8l2fu40jh0slnzjsku72w9/