AI 精选动态
智能评分 65
Meituan LongCat-2.0 发布
AI 推荐理由
Zero-Compute Experts 实现动态计算分配，MOPD 架构提供任务专用优化，性能与闭源模型相当，值得开发者探索高效推理方案。核心解读
Meituan_LongCat 发布 LongCat-2.0 模型，1.6T 参数（48B 活动），原生 1M 上下文，配备 LSA 稀疏注意力机制、Zero-Compute Experts 动态激活机制和 MOPD 三专家组路由。在 SWE-bench Pro 获得 59.5 分，与主流闭源模型性能相当。定价为每百万tokens输入/缓存/输出 0.015/0.75/2.95 美元，于 SiliconFlow 上线。
全文
The full model behind "Owl Alpha" on @OpenRouter is here🦉
Let's meet @Meituan_LongCat 's latest flagship model, LongCat-2.0
Now Day 0 live on SiliconFlow 🔥
💰 Input Cache/Input/Output: $ 0.015/0.75/2.95 per 1M tokens
⚙️ 1.6T-param MoE (~48B active) · Native 1M context window
🧠 Built for agentic coding from the ground up:
◆ LSA: sparse attention that scales efficiently to 1M
◆ Zero-Compute Experts: dynamic 33B–56B active/token, no wasted compute
◆ MOPD: three specialized expert groups (Agent / Reasoning / Interaction), gate-routed per task
🏆 59.5 SWE-bench Pro: performance on par with mainstream close-sourced models
Start building with 🐱👇
https://video.twimg.com/amplify_video/2071831110300917760/vid/avc1/1920x1080/bu8L1sZKGrzHmeQX.mp4?tag=28
SiliconFlow (@SiliconFlowAI): Go build with 🐱 on SiliconFlow⬇️
https://t.co/rD1INVfoCJ