LongCat-2.5-Preview is now live. 1.6T parameters. ~48B active. A 1M-token context window. Natively multimodal. Built to take on long-horizon tasks. From terminals and browsers to GUIs, spreadsheets, and design tools. Try it now: 🚀 API: 💬 Chat:

Great to welcome the @huggingface team to our Shanghai office! 🤗 We had inspiring conversations about LongCat’s future model development and open-source plans, as well as some of the latest features and technical details of Hugging Face Transformers. Hope you enjoy the LongCat keychains! Looking forward to exploring more ways to collaborate across the open-source ecosystem 🐱🚀

AI agents can now propose changes, run experiments, interpret feedback, and refine technical artifacts over many iterations. But does that make them autonomous researchers? We evaluated 7 frontier models on 36 AI R&D tasks that require sustained experimentation, covering 756 trajectories in total. Final scores only tell part of the story. We looked at how agents frame solutions, turn ideas into working implementations, retain progress, and recover from failure. We also used controlled comparisons to study experience reuse and the effect of different harness designs. Three findings stood out. 1️⃣ Strong optimization performance does not necessarily mean genuine innovation. Agents can formulate and implement practical solutions, but their strongest solutions mainly adapt or combine established techniques. Only 3 of the 252 solutions qualified as novel approaches. 2️⃣ Reliability separates current models more than peak performance. Many models can find a strong solution once. Reaching it consistently is what separates them. Similar final scores can also hide very different bottlenecks in solution framing, execution, and feedback control. 3️⃣ Experience can help or mislead, while harnesses mainly affect reliability. Accumulated experience usually improves the next solution by preserving useful discoveries, but it can also carry forward misleading conclusions or anchor agents to local optima. Harness choice mainly affects stability across repeated runs. Automated harness optimization produced gains that transferred to held out tasks and another model. Overall, current agents operate more like engineering optimizers than fully autonomous researchers. They can automate parts of the research loop, but reliable performance, effective experience reuse, and genuine novelty remain open challenges. 📄 Paper: 🌐 Project:

LongCat-2.0 is now live on TokenRhythm. 1.6T MoE • ~48B active • native 1M context • MIT open weights Trained end-to-end on domestic AI ASIC superpods and designed for repository-scale coding and agentic workflows. Ready to use with OpenSquilla. 🦐 #OpenSourceAI# #AIAgents#

.@Meituan_LongCat's LongCat-2.0 is now live on Infron 🐱 Built for agentic coding: ▸ 1.6T MoE, roughly 48B active ▸ 1M-token context, 128K max output ▸ Native tool calls, streaming, and a thinking mode you can switch off ▸ Speaks OpenAI and Anthropic formats, so Claude Code needs no shim ▸ $0.75 / $2.95 per M, on the same key as 400+ other models Try it:

🐱 LongCat-2.0 is now fully open-source — MIT licensed, no restrictions. Since our launch a few days ago, the response from the community has been incredible. Thank you for all the feedback, discussions, and interest. Today, we’re releasing the model weights and inference code to everyone. ◆ 1.6T MoE · ~48B active · 1M token context ◆ Agent-native: Integrates directly with Claude Code, OpenClaw, and Hermes Agent ◆ Deployment: Support both GPU and NPU platforms— verified on large-scale domestic clusters 📑 Tech Blog: 🤗 HuggingFace: 💻 GitHub: 🪄 ModelScope: 👇 Inference Code GPU: NPU:

Introducing LongCat-2.0 🐱 1.6T parameters · MoE with ~48B active · 1M context The full model behind Owl Alpha on @OpenRouter — now available. Built for agentic coding from the ground up: ◆ LongCat Sparse Attention (LSA) — scales efficiently for 1M-context tokens ◆ Zero-Compute Experts — dynamic activation 33B–56B per token, zero wasted compute ◆ MOPD — three specialized expert groups (Agent / Reasoning / Interaction), gate-routed per task How it stacks up: → Terminal-Bench 2.1: 70.8 → SWE-bench Pro: 59.5 (GPT-5.5: 58.6) → SWE-bench Multilingual: 77.3 → FORTE: 73.2 · RWSearch: 78.8 · BrowseComp: 79.9 📖 Tech Blog: Try it across different scenarios 🧵👇