AI Daily · November 2026 · Monday
AITOPNEWS
Meituan releases LongCat-2.5, a 1.6-trillion-parameter agent model trained entirely on Chinese chips
Meituan released the weights of LongCat-2.5, a 1.6-trillion-parameter mixture-of-experts model built for autonomous multi-step agent and coding tasks, trained entirely on Chinese hardware with no Nvidia GPUs. The model activates about 48 billion parameters per token, handles up to one million tokens via API, understands text, images and video, and is available under the MIT license with a self-hosting guide for a 194-chip Huawei Ascend Atlas A2 cluster. Meituan reports 88.4 on Terminal Bench v2.1 versus Claude Opus 5's 89.1 and GPT-5.6 Sol's 88.8, and 38.5 on the harder Terminal Bench v4.0 against Opus 5's 53.9.
Model releases
World Labs releases Atlas, an omni world model for spatial intelligence
World Labs introduced Atlas, a next-generation world model pretrained from scratch as a multimodal autoregressive diffusion transformer that natively operates on text, images, video and 3D, with all inputs grounded in a shared spatial context.
Anthropic's Claude Fable 5.2 paints images stroke by stroke with simulated brush physics
2 creates paintings by writing Python programs of 1,000 to 2,000 lines that drive a simulated paint engine, placing hundreds of thousands of individual brushstrokes on a virtual canvas instead of using diffusion. Anthropic developer Jake Eaton showed a photorealistic parking garage scene built from roughly 700,000 simulated strokes, with no external image model and no pre-rendered or traced source images.