MiniMax-M2: A New Open-Source LLM Champion Emerges,Excelling in Agentic Tool Use
Table of Contents
Watch out,DeepSeek and qwen! There’s a new king of open source large language models (LLMs),especially when it comes to something enterprises are increasingly valuing: agentic tool use – that is,the ability to go off and use other software capabilities like web search or bespoke applications – without much human guidance.
That model is none other than MiniMax-M2 the latest LLM from the Chinese startup of the same name. And in a big win for enterprises globally, the model is available under a permissive, enterprise-kind MIT License, meaning it is made available freely for developers to take, deploy, retrain, and use how they see fit – even for commercial purposes. It can be found on Hugging Face, GitHub and ModelScope as well as through MiniMax’s API here. It supports OpenAI and Anthropic API standards, as well, making it easy for customers of said proprietary AI startups to shift out their models to MiniMax’s API, if they want.
according to independant evaluations by Artificial Analysis a third-party generative AI model benchmarking and research institution, M2 now ranks first among all open-weight systems worldwide on the Intelligence Index-a composite measure of reasoning, coding, and task-execution performance.
In agentic benchmarks that measure how well a model can plan, execute, and use external tools-skills that power coding assistants and autonomous agents-MiniMax’s own reported results, following the Artificial Analysis methodology, show τ-Bench 77.2, BrowseComp 44.0, and FinSearchComp-global 65.5.
These scores place it at or near the level of top proprietary systems like GPT-5 (thinking) and Claude Sonnet 4.5,making MiniMax-M2 the highest-performing open model yet released for real-world agentic and tool-calling tasks.
What It Means For Enterprises and the AI race
Built around an efficient Mixture-of-Experts (MoE) architecture, MiniMax-M2 delivers high-end capability for agentic and developer workflows while remaining practical for enterprise deployment.
For technical decision-makers, the release marks an important turning point for open models in business settings. MiniMax-M2 combines frontier-level reasoning with a manageable activation footprint-just 10 billion active parameters out of 230 billion total.
This design enables enterprises to operate advanced reasoning and automation workloads on fewer GPUs, achieving near-state-of-the-art results without the infrastructure demands or licensing costs associated with proprietary frontier systems.
Artificial Analysis’ data show that MiniMax-M2’s strengths go beyond raw intelligence scores. The model leads or closely trails top proprietary systems such as GPT-5 (thinking) and Claude Sonnet 4.5 across benchmarks for end-to-end coding, reasoning, and agentic tool use.
Its performance in τ-Bench,SWE-Bench,and BrowseComp indicates particular advantages for organizations that depend on AI systems capable of planning,executing,and verifying complex workflows-key functions for agentic and developer tools inside enterprise environments.
As LLM engineer Pierre-Carl Langlais aka Alexander Doria posted on X: “MiniMax [is] making a case for mastering the technology end-to-end to get actual agentic automation.”
Compact design,Scalable Performance
MiniMax-M2’s technical architecture is
## MiniMax Launches M2,a Powerful Open-Source Model Challenging GPT-5
MiniMax,a leading AI model provider,has announced the release of MiniMax-M2,a new open-source large language model (LLM) designed to rival the capabilities of OpenAI’s GPT-5. The model is gaining attention for its strong reasoning abilities, cost-effectiveness, and flexible deployment options.
###Notable Performance and Capabilities
MiniMax-M2 is a 72B parameter model built on a Mixture-of-Experts (MoE) architecture.According to MiniMax, M2 achieves state-of-the-art performance on several key benchmarks, including MMLU, HellaSwag, and TruthfulQA.
A key feature of M2 is its advanced tool-calling functionality. This allows the model to interact with
external tools and APIs via structured XML-style calls. This functionality allows MiniMax-M2 to serve as the reasoning core for larger agent frameworks, executing dynamic tasks such as search, retrieval, and computation through external functions.###Open Source Access and Enterprise Deployment Options
Enterprises can access the model through the MiniMax Open Platform API and MiniMax Agent interface (a web chat similar to ChatGPT), both currently free for a limited time.
MiniMax recommends SGLang and vLLM for efficient serving, each offering day-one support for the model’s unique interleaved reasoning and tool-calling structure.
Deployment guides and parameter configurations are available through MiniMax’s documentation.
###Cost Efficiency and Token Economics
As Artificial Analysis noted, MiniMax’s API pricing is set at $0.30 per million input tokens and $1.20 per million output tokensamong the most competitive in the open-model ecosystem.
|
Provider |
Model (doc link) |
Input $/1M |
Output $/1M |
Notes |
|
MiniMax |
$0.30 |
$1.20 |
Listed under “Chat Completion v2” for M2. |
|
|
OpenAI |
$1.25 |
$10.00 |
Flagship model pricing on OpenAI’s API pricing page. |
|
|
OpenAI |
$0.25 |
$2.00 |
Cheaper tier for well-defined tasks. |
|
|
Anthropic |
$3.00 |
$15.00 |
Anthropic’s current per-MTok list MiniMaxAI Launches MiniMax-M2: A New Open-Weight AI Model for Enterprise ApplicationsMiniMaxAI has released MiniMax-M2, a new open-weight large language model (LLM) designed to address the complex needs of engineering teams and data-intensive applications. This release positions MiniMaxAI as a significant player in the rapidly evolving landscape of open-source AI, offering a powerful choice to closed-source models with features like structured function calling, long-context retention, and efficient attention mechanisms – all without the constraints of vendor lock-in or restrictive compliance requirements. MiniMax-M2 is notable for prioritizing practical utility and controllable reasoning over sheer model size,making it a highly enterprise-ready solution. Understanding Open-Weight AI and its BenefitsOpen-weight AI models, like MiniMax-M2, differ from closed-source models (like those from OpenAI or Google) in a crucial way: the model weights – the core parameters that define its intelligence – are publicly available. This openness offers several key advantages: * clarity & Auditability: Organizations can inspect the model’s inner workings, ensuring it aligns with their security and ethical standards. MiniMax-M2: Key features and CapabilitiesMiniMax-M2 is engineered to excel in complex reasoning tasks and data processing pipelines. Here’s a breakdown of its core features: * Structured Function Calling: This allows the model to interact with external tools and APIs in a predictable and reliable manner, enabling automation of tasks and integration with existing systems. https://www.minimaxir.com/blog/minimax-m2 * Long-Context Retention: MiniMax-M2 can process and remember information from significantly longer inputs than many other models, crucial for tasks requiring understanding of extensive documentation or complex conversations. MiniMaxAI and the Rise of Chinese Open-Weight AIThe launch of MiniMax-M2 underscores the growing prominence of Chinese AI research groups in the open-weight model space.Following contributions from companies like DeepSeek, Alibaba (with its Qwen series), and Moonshot AI, MiniMaxAI is contributing to a trend toward accessible, efficient, and practical AI systems. According to artificial Analysis, MiniMax-M2 represents a shift in focus towards agentic capability and reinforcement learning, prioritizing controllable reasoning and real-world application over simply increasing model size. https://artificialanalysis.co/minimax-m2-review/ Enterprise Readiness and Practical Applicationsminimax-M2 is designed with enterprise needs in mind. Its open licensing allows for internal deployment with full transparency, enabling organizations to audit, fine-tune, and customize the model to their specific requirements. This makes it a viable foundation for building intelligent systems that can: * Automate complex workflows: Leverage structured function calling to integrate with existing tools and APIs. Key Takeaways* Open-Weight Advantage: MiniMax-M2 offers the benefits of transparency, customization, and freedom from vendor lock-in. looking AheadMiniMaxAI’s continued expansion of its open-weight model lineup signals a commitment to driving innovation in the AI space. As the demand for accessible, customizable, and obvious AI solutions grows, MiniMax
previous post
2025 World Series Game 4: Storylines & What to Watch |