๐ค China's 1.6T-Param LongCat-2.0 Just Outscored GPT-5.5 on Coding โ and It's Free
Meituan just open-sourced LongCat-2.0, a 1.6 trillion-parameter Mixture-of-Experts model built on a cluster of 50,000+ domestic Chinese ASICs โ zero Nvidia GPUs required. It scored 59.5 on SWE-bench Pro, edging out GPT-5.5's 58.6, and ships with a 1-million-token context window under an MIT license. This thing was trained entirely on export-controlled-free hardware, which is kind of a big deal. Would you swap your paid coding AI for this?
โก 31 buzzing
306 views
๐ 0 subscribed