Mistral AI's Cascade Distillation Empowers Small Models with Large Model Capabilities▲ 92 Deloitte and Nvidia Expand Partnership for Industrial AI Solutions▲ 90 New Study Reveals AI's Ability to Expose Hidden Online Identities▲ 90 Intel Advances 6G Strategy with Foundry and AI Partnerships▲ 88 Liverpool FC Files Complaint Against X Over Grok AI-Generated 'Despicable' Tweets▲ 85 Sarvam AI Releases Open-Weight Models, Benchmarked Against DeepSeek and Gemini▲ 82 Open-Source Coding Agents Streamlining Developer Workflows▶ 80 Emerging Trend: AI for Emotional Processing and Mental Anguish Release▶ 78 New Tool 'llmfit' Recommends Optimal AI Models Based on System Hardware▶ 68 Google Releases Open-Source CLI for Workspace Management▶ 60///Mistral AI's Cascade Distillation Empowers Small Models with Large Model Capabilities▲ 92 Deloitte and Nvidia Expand Partnership for Industrial AI Solutions▲ 90 New Study Reveals AI's Ability to Expose Hidden Online Identities▲ 90 Intel Advances 6G Strategy with Foundry and AI Partnerships▲ 88 Liverpool FC Files Complaint Against X Over Grok AI-Generated 'Despicable' Tweets▲ 85 Sarvam AI Releases Open-Weight Models, Benchmarked Against DeepSeek and Gemini▲ 82 Open-Source Coding Agents Streamlining Developer Workflows▶ 80 Emerging Trend: AI for Emotional Processing and Mental Anguish Release▶ 78 New Tool 'llmfit' Recommends Optimal AI Models Based on System Hardware▶ 68 Google Releases Open-Source CLI for Workspace Management▶ 60

← Back to Briefing

China's MiniMax Launches Cost-Effective AI Model M2.5, Challenging Global Leaders

Importance: 92/1007 Sources

Why It Matters

This release signifies increasing competition in the global AI market, with Chinese firms offering powerful models at disruptive price points, potentially accelerating AI adoption and innovation worldwide.

Key Intelligence

■Chinese AI company MiniMax has released its new flagship model, M2.5.
■The M2.5 model claims performance comparable to leading global AI models such as GPT-5, Claude Opus, and Gemini.
■A key differentiator is its significantly lower operational cost, reportedly 1/20th the price of GPT-5 and around $1 per hour to run.
■MiniMax positions M2.5 as designed for "real-world productivity" and aims to compete directly with OpenAI and Anthropic.
■The announcement has fueled optimism in the Chinese AI sector, leading to a rally in related stocks.

Source Coverage

Google News - AI

MiniMax released the M2.5 model: costing 1 dollar to run for 1 hour, with a price only 1/20 of GPT-5, and performance comparable to Claude Opus. - 富途牛牛

Google News - AI & Models

China’s MiniMax releases AI model ‘designed for real-world productivity’ - South China Morning Post

Google News - AI & Models

Chinese AI stocks rally on optimism over new model releases - Investing.com

Google News - AI & Models

Analysis: China's AI models and chips align on day one - digitimes

Google News - AI & Models

China’s MiniMax launches cheap AI model to rival Anthropic, OpenAI - Tech in Asia

Google News - AI & Models

China’s MiniMax Launches New Flagship AI Model M2.5 - The Information

Google News - AI

Chinese AI company MiniMax announces 'M2.5,' which has the same performance as GPT, Gemini, and Claude but costs less than one-tenth of the cost - GIGAZINE