Veriwire
← Back to the wire

Meituan releases a 1.6-trillion-parameter agent model that never touched an Nvidia GPU

AchievementModelNov 15, 2026

Meituan has released the weights of LongCat-2.5, a 1.6-trillion-parameter agent model trained entirely on Chinese hardware without Nvidia GPUs. The mixture-of-experts model activates roughly 48 billion parameters per token, handles up to one million input tokens, and adds image understanding built on LongCat-2.0 and research from LongCat-Next. It scores 88.4 on Terminal Bench v2.1, slightly below Claude Opus 5's 89.1, and is available under the MIT license.

Why it matters

Meituan, the Chinese food delivery company, has publicly released the weights of LongCat-2.5, a mixture-of-experts model designed for autonomous agents working through long, multi-step coding tasks. The release is notable because the model was trained entirely on Chinese hardware, with no Nvidia GPUs involved, and its self-hosting guide describes a cluster built around Huawei Ascend accelerators. Meituan quantizes the model to a low-bit precision that those Chinese chips process natively, keeping accuracy close to the uncompressed version while boosting throughput. On agent benchmarks, LongCat-2.5 lands just behind top Western models like Claude Opus 5 on terminal tasks, though Meituan ran some evaluations in its own environment. The weights ship under a permissive MIT license, allowing commercial use, modification, and redistribution. The move signals that open-weight Chinese models trained on domestic hardware are closing the gap with leading US systems on common benchmarks.

Receipt № 28961 source · awaiting confirmation ◐

Evidence

1source· awaiting independent confirmation

No score is assigned. Sources and their independence are shown in the citation chain below.

Citation chain · 1 source

01med
the-decoder.comBloghttps://the-decoder.com/meituan-releases-a-1-6-trillion-parameter-agent-model-that-never-touched-an-nvidia-gpu/The Decoder details LongCat-2.5's training without Nvidia GPUs, its low-bit quantization for Huawei Ascend chips, and its benchmark scores relative to leading Western models.
PRIMARY
NvidiaCompanyClaude Opus 5ModelMeituanCompanyLongCat-2.5ModelLongCat-2.0ModelLongCat-NextModel

Related

Nov 15Bland Launches Tofu-1, a Full-Duplex Speech-to-Speech Model for CallsNov 15EUROPA, the EU's open AI model, is aiming for nearly 1 trillion parametersNov 14Anthropic can already retrain Mythos with "a fraction of the compute originally needed," says former researcher Coxon
Canonical: https://the-decoder.com/meituan-releases-a-1-6-trillion-parameter-agent-model-that-never-touched-an-nvidia-gpu/