Meituan's LongCat team released the model weights for LongCat-2.5 under the MIT license, marking a significant milestone as the 1.6 trillion-parameter mixture-of-experts (MoE) model was trained entirely on Chinese domestic hardware. The model activates approximately 48 billion parameters per token and supports a 1-million-token context window.
Benchmark results show LongCat-2.5 scoring 88.4 on Terminal Bench v2.1, placing it just behind Claude Opus 5 (89.1) and GPT-5.6 Sol (88.8). Meituan also published a 4-bit quantized version with a self-hosting guide built on 194 Huawei Ascend chips — notably offering no NVIDIA GPU option. Limited-time API pricing stands at roughly €0.27 per million input tokens and €1.07 per million output tokens.
Meituan Open-Sources 1.6T-Parameter LongCat-2.5 Model Trained Entirely on Domestic Chinese Hardware
Avertissement : Le contenu proposé sur Phemex News est à titre informatif uniquement. Nous ne garantissons pas la qualité, l'exactitude ou l'exhaustivité des informations provenant d'articles tiers. Ce contenu ne constitue pas un conseil financier ou d'investissement. Nous vous recommandons vivement d'effectuer vos propres recherches et de consulter un conseiller financier qualifié avant toute décision d'investissement.