# Meituan releases LongCat-2.5 weights under MIT License for coding and agents

> Original title: Meituan has released the weights for LongCat-2.5 under the MIT License. The native multimodal MoE is built for long-horizon coding and ag...

- Source: ModelScope
- Published: 2026-11-10T08:00:06.000Z
- Bad Influence: https://badinfluence.news/items/i0ugsjwlvovu1szt4ge9tdkv8
- Original: https://x.com/ModelScope2022/status/2120063938448515358

## Summary

Meituan has released the weights for LongCat-2.5 under the MIT License. The native multimodal MoE has 1.6T total parameters, with about 48B activated per token, and is built for long-horizon coding and agent tasks. In Meituan's tests, it improves substantially over LongCat-2.0 on DeepSWE and Terminal-Bench v4.0, and it leads in the 3DCodeBench Text-to-3D comparison.

## Why it matters

The release combines open MIT weights with benchmark gains over LongCat-2.0 in coding and agent tasks, showing where Meituan's model is claimed to lead.

## Post

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://modelscope.ai/models/meituan-longcat/LongCat-2.5

• 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.

![](https://badinfluence.news/api/img-proxy?u=https%3A%2F%2Fpbs.twimg.com%2Fmedia%2FHWv66nuXoAALqWc.jpg&mode=thumb&exp=1794528000&sig=279f21d272a18c92)