88Z AI, 13 April — MiniMax has officially open-sourced MiniMax M2.7, a new self-evolving AI model deeply iterated by AI itself. It excels in building complex Agent systems and performs well in software engineering and office scenarios. On its first day of release, it achieved compatibility with major domestic and international chip and inference platforms.
At a glance
- MiniMax M2.7 is the first open-source model deeply iterated by AI, excelling in building complex Agent teams.
- It achieved compatibility with major domestic and international chips and inference platforms like Huawei Ascend and NVIDIA on day one.
The update
MiniMax has officially launched and open-sourced its new generation artificial intelligence model, MiniMax M2.7. The core breakthrough of this model lies in its self-evolution mechanism, where AI is deeply involved in the iteration process, significantly enhancing intelligence levels. In terms of functional applications, M2.7 supports the construction of complex Agent Harnesses and Agent Teams, demonstrating powerful multi-agent collaboration capabilities. In practical tests, the model has shown excellent performance in software engineering development and daily office scenarios. Regarding ecosystem compatibility, MiniMax M2.7 achieved broad hardware and platform compatibility on its first day of release. It has completed adaptation with major domestic and international chip manufacturers, including Huawei Ascend, Moore Threads, MetaX, Kunlunxin, and NVIDIA. Additionally, the model has been integrated with well-known inference platforms such as Together AI and Fireworks, providing developers with diverse deployment options and ensuring efficient operation across different computing infrastructures.
Sources & editorial note
This article is an AI-assisted synthesis by 88Z AI based on the public sources below. Product claims and availability are attributed to their publishers. It does not represent independent testing or first-hand reporting.
- MiniMax 稀宇科技Source published
Source dates above refer to the original announcements; the date under the headline is this article’s publication on 88Z AI.