option
Home
List of Al models
MiniMax-M2.7
Model parameter quantity
230B
Model parameter quantity
Affiliated organization
MiniMax
Affiliated organization
Open Source
License Type
Release time
March 18, 2026
Release time
Model Introduction
MiniMax-M2.7 is MiniMax’s new-generation general-purpose large model. By building an Agent Harness system, it deeply participates in its own training and optimization process, can take on 30%–50% of the workload in some R&D scenarios, and achieves about a 30% improvement on internal evaluation benchmarks.
Swipe left and right to view more
Language comprehension ability Language comprehension ability
Language comprehension ability
Often makes semantic misjudgments, leading to obvious logical disconnects in responses.
6.8
Knowledge coverage scope Knowledge coverage scope
Knowledge coverage scope
Possesses core knowledge of mainstream disciplines, but has limited coverage of cutting-edge interdisciplinary fields.
8.2
Reasoning ability Reasoning ability
Reasoning ability
Unable to maintain coherent reasoning chains, often causing inverted causality or miscalculations.
5.2
Related model
MiniMax-M2.1 The latest flagship model released by Minimax features excellent general capabilities, advanced programming capabilities, powerful agent performance, and efficient architecture design.
MiniMax-M2 The latest flagship model released by Minimax features excellent general capabilities, advanced programming capabilities, powerful agent performance, and efficient architecture design.
MiniMax-Text-01 MiniMax-Text-01 is a powerful language model with 456 billion total parameters, of which 45.9 billion are activated per token. To better unlock the long context capabilities of the model, MiniMax-Text-01 adopts a hybrid architecture that combines Lightning Attention, Softmax Attention and Mixture-of-Experts (MoE).
MiniMax-M1-80k The world's first open-weight, large-scale hybrid-attention reasoning model released by Minimax
abab6.5 abab6.5 is an API model produced by MiniMax, with the version number being abab6.5. The abab6.5 serie is a trillion-parameter Mixture of Experts (MoE) large language model. The abab6.5 is suitable for complex scenarios, such as application problem calculations, scientific computations, and other similar scenarios. The abab6.5s is suitable for general scenarios.
Relevant documents
Cloudflare Unveils Open Source AI Agent Workspace, Empowering Employees to Build Custom Workflows Without Code Cloudflare has officially released Cloudflare OS, an open-source AI agent workspace that operates directly in the browser. This platform integrates a wide array of customizable, shareable micro-applications. Far from being a mere wrapper around gener
Musk’s AI Credit Card Checkout Sparks Backlash After OpenAI Exec’s Mockery Elon Musk recently unveiled Grok’s new shopping capability on social media, explaining that users can snap a photo of their credit card and send it to the chat interface. The AI will then scour the web for the best prices and finalize the purchase on
How to optimize SEO in 2024 for Google and Yahoo Japan Master Storyboard Drawing: Expert Tips for Non-artistsTable of Contents:IntroductionThe Importance of StoryboardingCommon Misconceptions about StoryboardingThe Golden Rule for a Great StoryboardDifferent Approaches to Storyboarding 5.1 Visual Approac
Zhiyuan Robotics Unveils Self-Developed Embodied Intelligence During Product Release Week AGIBOT, a pioneer in embodied intelligence, announced on April 3rd that it will host a six-day "AGIBOT AI Week" beginning next Monday, April 7th. This event will highlight a series of core proprietary breakthroughs in embodied intelligence with unpre
Neptune Medical Gets FDA Clearance for Triton 1 Robotic System Triton is a gastrointestinal robotic endoscopy system engineered for intuitive navigation, precision, and control across the entire colon. | Source: Neptune MedicalNeptune Medical Inc. announced this week that it has received FDA clearance for its Tr
Model comparison
Start the comparison
OR