MiniMax M3 Debuts, Eclipsing GPT-5.5 and Gemini 3.1 Pro on Key Benchmark Performance for Just 5-10% of the Cost
The Shanghai-based startup says the model cuts compute needs to one-twentieth of prior levels and will publish weights on HuggingFace and GitHub within 10 days.
- On Sunday, Chinese AI startup MiniMax released its M3 large language model, featuring a 1-million-token context window and native multimodality paired with frontier-tier coding performance at reduced costs.
- Redesigned architecture using sparse attention reduces computational requirements to 1/20th of previous levels, slashing inference costs. Standard Transformer networks scale quadratically, making long inputs prohibitively expensive.
- The model records 59.0% on SWE-Bench Pro, outperforming Gemini 3.1 Pro, while trailing Claude Opus 4.8 on code modification tasks. M3 maintains competitive execution against DeepSeek-V4 Pro Max on agentic categories.
- MiniMax plans to release M3 under an open-weights license within 10 days, enabling enterprise users to run the model locally. API pricing starts at $0.3 per million input tokens, undercutting proprietary U.S. competitors.
- Developers confirmed day-one compatibility with tools like Cline, the agentic AI coding harness. In a 12-hour autonomous test, M3 produced 18 commits and 23 experimental figures, demonstrating capability for complex software engineering projects.
13 Articles
13 Articles
MiniMax M3 debuts, eclipsing GPT-5.5 and Gemini 3.1 Pro on key benchmark performance for just 5-10% of the cost
Big news in enterprise AI broke over the weekend as Chinese AI startup MiniMax released its highly anticipated M3 large language model on Sunday evening Eastern time, pairing frontier-tier coding and agentic performance with a 1-million-token context window and native multimodality for a fraction of the cost of leading proprietary models, with pricing starting at just $20 per month under its new subscription token plans. The company's leadership…
MiniMax M3 Decodes 1M Tokens 15x Faster — and It Shouldn't Be This Cheap
Last Updated on June 3, 2026 by Editorial Team Author(s): Chew Loong Nian – AI ENGINEER Originally published on Towards AI. MiniMax M3 Decodes 1M Tokens 15x Faster — and It Shouldn’t Be This Cheap On June 1, a Shanghai lab quietly shipped a model that decodes a 1-million-token context 15.6x faster than its own previous generation — and charges you roughly 8% of what Claude Opus costs to do it. I spent two days poking at MiniMax M3 through the AP…
Open Source MiniMax M3 Outperforms Opus 4.7 for a Fraction of the Cost
The MiniMax M3 is making waves in the AI community as an open source model that combines advanced capabilities with affordability. Highlighted by World of AI, this model stands out for its ability to handle both text and visual data through multimodal reasoning, allowing applications like image captioning and multimedia generation. Its support for a […] The post Open Source MiniMax M3 Outperforms Opus 4.7 for a Fraction of the Cost appeared firs…
MiniMax, an AI development company based in Shanghai, China, has announced its AI model, "MiniMax M3." MiniMax M3 has achieved benchmark scores that compete with GPT-5.5 and Claude Opus 4.7, and the company has stated that it will be distributed as an open model within 10 days of its announcement.
MiniMax M3 AI Model New Open Alternative Shakes Up GPT-5.5
MiniMax M3 AI model, launched on June 1, 2026, marks a seismic shift in the artificial intelligence landscape. This deep research report breaks down the architectural breakthroughs, empirical benchmarks, and broader industry implications of a model designed to challenge closed-source titans like OpenAI’s GPT-5.5 and Google’s Gemini 3.1 Pro. MiniMax introduced M3 as its flagship artificial intelligence model, specifically targeting expansion into…
Coverage Details
Bias Distribution
- 100% of the sources are Center
Factuality
To view factuality data please Upgrade to Premium
