Meituan Releases New AI Model Trained Entirely on Chinese Chips

2026-07-01 · Source: The Information · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Emerging Technologies & Innovation · Depth: Novice, quick

Summary

Chinese food delivery giant Meituan has unveiled LongCat-2.0, a new open-source large language model featuring 1.6 trillion parameters. This model is notable for being trained entirely on Chinese-made chips, marking a significant development in domestic AI infrastructure and technological self-reliance. Meituan, which previously developed its first large language model last year, positions LongCat-2.0 as a key advancement in its AI capabilities. The release underscores the company's commitment to fostering self-sufficiency in advanced AI development and highlights broader trends within China to reduce reliance on foreign hardware for critical technological advancements.

Key takeaway

For Directors of AI/ML evaluating supply chain resilience, Meituan's LongCat-2.0 demonstrates the viability of training large language models entirely on domestic hardware. You should consider how this trend impacts your long-term hardware procurement strategies and potential geopolitical risks. This development suggests increasing options for AI development independent of traditional global chip suppliers, potentially diversifying your infrastructure choices.

Key insights

Meituan's LongCat-2.0, trained on Chinese chips, signals growing domestic AI hardware independence.

Principles

• Domestic chip training enables AI self-sufficiency.
• Open-source models foster broader adoption.
• Large parameter counts drive model capability.

In practice

• Evaluate LongCat-2.0 for Chinese-chip-compatible AI.
• Explore open-source LLMs for specific applications.

Topics

• Large Language Models
• Meituan
• Chinese Chips
• AI Infrastructure
• Open-source AI
• LongCat-2.0

Best for: NLP Engineer, Research Scientist, AI Scientist, Director of AI/ML, Tech Journalist

Related on AIssential

• Meituan open sources LongCat-2.0, the 1.6T, near-frontier agentic coding model that's been leading OpenRouter — trained entirely on Chinese chips — Meituan's LongCat-2.0 demonstrates near-frontier AI can be developed on domestic Chinese ASICs with competitive performance and cost.
• Paper Digest: ICLR 2026 Papers & Highlights — The ICLR 2026 highlights reveal a strong focus on enhancing LLM reasoning, multi-modal capabilities, and robust AI agent development.
• Fable 5 IS BACK? GPT 5.6 & Gemini 3.5 Pro Delayed, NEW OpenAI AI Chip, & QWEN STEALING! AI NEWS — AI model capabilities can pose national security risks and drive industrial-scale espionage.
• Meituan's LongCat-2.0 shows China can train massive AI models without Nvidia — China has demonstrated the ability to train trillion-parameter AI models using only domestic hardware, challenging US export controls.
• Marc Andreessen introspects on The Death of the Browser, Pi + OpenClaw, and Why "This Time Is Different" — Current AI advancements are an "80-year overnight success" built on foundational research, now demonstrating real-world functionality.
• not much happened today — AI advancements focus on harder benchmarks, agentic systems, optimized training, specialized inference, and critical supply-chain security.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by The Information.