美團以 MIT License 釋出原生多模態 MoE 模型 LongCat-2.5 權重,模型總引數量為 1.6T,每個 token 約啟用 48B 引數,面向長程式開發與 AI 助理任務。
AI 伴隨問答
針對這篇報導有疑問?可直接點選快捷問題或向 AI 追問背景脈絡與技術細節:
LongCat-2.5 的比較涵蓋程式開發、視覺程式設計與 3D 建模,也列出 NPU 上 INT4 推論的吞吐量變化,呈現其任務範圍與測試結果。
Meituan has released the weights for LongCat-2.5 under the MIT License. The native multimodal MoE is built for long-horizon coding and agent tasks, with 1.6T total parameters and about 48B activated per token.
🤖 https://t.co/l5my9potsI
• Stronger coding agents. LongCat-2.5 improves substantially over LongCat-2.0 on DeepSWE and Terminal-Bench v4.0 in Meituan’s tests.
• Visual coding. The model processes image and text tokens together, so it can use a screenshot as a reference when writing webpage code.
• 3D modeling through code. The model can turn a text description into a Blender script and leads the models in Meituan’s 3DCodeBench Text-to-3D comparison.
• Training for longer tasks. Agentic RL uses real engineering tasks and reproducible agent failures. LongCat-2.5 also posts the highest BrowseComp score in Meituan’s reported comparison.
• INT4 inference on NPUs. With quantized KV cache, W4A4 nearly doubles decode throughput over W8A8KV16 in Meituan’s stress test.