Today for AI
HOT RADAR
llmHEAT 11.4°

AI CURATED HOT STORY · 11/2/2026

Meituan Releases LongCat-2.5: A 1.6T Parameter Native Multimodal Model for Long-Horizon Agents

1 reports archived1 independent sourcesupdated 11/2/2026, 09:18:38
Synthesis & Latest Updates
1 Sources Cross-Validated

Meituan's LongCat team has released LongCat-2.5, a native multimodal large model with 1.6 trillion total parameters. By leveraging large-scale long-horizon reinforcement learning and multi-level discrete visual tokens, the model significantly enhances end-to-end engineering capabilities in complex codebase development, Visual Coding, and long-horizon agent tasks. Online access and APIs are currently available, with open-source weights coming soon.

LATEST/Meituan's LongCat team has released LongCat-2.5, a native multimodal large model with 1.6 trillion total parameters. By leveraging large-scale long-horizon reinforcement learning and multi-level discrete visual tokens, the model significantly enhances end-to-end engineering capabilities in complex codebase development, Visual Coding, and long-horizon agent tasks. Online access and APIs are currently available, with open-source weights coming soon.

Hourly heat curve

2 fully observed hours
Fewer than 3 fully observed hours; trend not plotted yet.

Coverage timeline

Total 1 reports · Latest first
  1. IT之家 · 智能时代T2·85 pts

    Meituan Releases LongCat-2.5: A 1.6T Parameter Native Multimodal Model for Long-Horizon Agents

    • LongCat-2.5 features 1.6T total parameters with ~48B active per token, indicating a massive MoE architecture.
    • It incorporates large-scale long-horizon reinforcement learning to optimize end-to-end capabilities for complex codebase construction and full project development.
    • Visual signals are represented as multi-level discrete tokens for unified text-vision processing, significantly boosting Visual Coding performance.