Back to News
MiniMaxJune 1, 2026

MiniMax M3

Featured on Blog
Paid
MiniMax
Open Weight
LLM
Multimodal
Long Context
Coding
AI Model
Frontier Model

Explore MiniMax M3

Visit the official website to learn more and get started

Featured
MiniMax M3: The First Open-Weight Model with Frontier Coding, 1M Context, and Native Multimodality

🔥 This release made it to our blog

MiniMax M3: The First Open-Weight Model with Frontier Coding, 1M Context, and Native Multimodality

MiniMax has released M3, the first open-weight frontier model to combine top-tier coding performance, a 1 million token context window via the new MSA sparse attention architecture, and native multimodality — all in a single model that can also control a desktop computer.

### TL;DR

MiniMax M3 is the first open-weight frontier model to combine frontier-level coding, 1M token context via MSA (MiniMax Sparse Attention), and native multimodality (image + video input + desktop control) in a single model. It surpasses GPT-5.5 and Gemini 3.1 Pro on SWE-Bench Pro, beats Opus 4.7 on SVG-Bench, and leads on Claw-Eval for autonomous agents.

Key Insights & Metrics

Pricing
$0.60 / million input token & $2.40 / million output token
Cost structure
Version
M3
Current release version
Hardware
GPU required for self-hosting; available via API (no hardware needed)
Compute requirements
Category
Paid
Licensing model
Region
China
Primary region

Key Features

  • MSA (MiniMax Sparse Attention) enables true 1M token context — a new sparse attention architecture that partitions KV into blocks more precisely than DSA/MoBA; per-token compute at 1M context is 1/20 of the previous generation, with 9× prefill speedup and 15× decode speedup versus full attention, while matching full attention on nearly all capability benchmarks
  • Frontier coding and agentic performance — SWE-Bench Pro: 59.0%, Terminal-Bench 2.1: 66.0%, MCP Atlas: 74.2%, KernelBench Hard: 28.8%; demonstrated 24-hour autonomous CUDA kernel optimization improving Hopper FP8 hardware utilization from 7.6% to 71.3% (9.4× speedup) across 147 benchmark submissions with zero human intervention
  • First open-weight model with all three frontier capabilities — native multimodality (image/video input, desktop computer control) trained from Step 0 with interleaved text+image data scaled to 100T tokens; independently reproduced an ICLR 2025 Outstanding Paper over 12 hours, 18 commits, and 23 experimental figures; available via MiniMax Code, Token Plan, and API

Related Releases

MiroThinker

MiroThinker is an open-source search agent model developed by MiroMindAI, designed for tool-augmented reasoning and real-world information seeking. It aims to match the deep research capabilities of leading AI models like OpenAI's Deep Research and Google's Gemini Deep Research.

MiroMindAI
Open

Letta Code SDK

The Letta Code SDK is a software development kit that enables developers to build deeply personalized agents with persistent memory that learn over time. It serves as the interface to Letta Code, facilitating the creation of stateful agents capable of continuous learning and improvement.

Letta AIJan 28
Open

OB-1

OB-1 is a self-improving coding agent developed by OpenBlock Labs, designed to autonomously handle the full development lifecycle, from project management to pull requests. It integrates seamlessly into existing workflows, enhancing productivity and code quality.

OpenBlock LabsSep 1
Open

Protenix

Protenix is an open-source, trainable PyTorch implementation of AlphaFold 3, designed for high-accuracy biomolecular structure prediction. It aims to advance accessible and extensible research tools for the computational biology community. ([github.com](https://github.com/bytedance/Protenix?utm_source=openai))

ByteDanceNov 5
Open

Discussion

0
Upvotes
0
Downvotes
0 reviews

Sign in to leave a review

Reviews

No reviews yet. Be the first to review!

🚀 Join the AI dev community — follow us everywhere

© 2026 MARKTECHPOST AI MEDIA INC. All rights reserved.Terms & ConditionsPrivacy Policy
Beta Mode