MiniMax M3
MiniMax M3 from MiniMax is an open-weight model with 1M context, native multimodality, agentic reasoning, and coding-focused workflows for developers.
Model Overview
Capabilities, design details, and architectural traits
MiniMax M3 - Coding & Agentic Frontier
MiniMax M3 is defined by three linked traits: coding, agentic reasoning, and native multimodality. It is presented as an open-weight model built to handle long-context work and multimodal input in one system.
| Distinctive trait | What it means |
|---|---|
| Coding & agentic focus | It is described as a model for agentic reasoning, tool use, coding, and long-context tasks. |
| 1M-context MSA | It uses MiniMax Sparse Attention to support up to 1M tokens of context. |
| Native multimodality | It is trained from step zero for text, image, and video understanding. |
M3 workflow
The model is described as using autonomous task decomposition, tool invocation, and multi-step reasoning. That makes its documented workflow centered on long-horizon execution rather than short single-turn answers.
Long-context use
M3 is also described as supporting long-range agent tasks, long-range coding, and long-video understanding. Its official positioning ties these together as one combined model design, not separate modes.
Benchmark Performance
Independent evaluations · Artificial Analysis
Accuracy & Capability Details
Compare Models Side-by-Side
Evaluate specifications, pricing, and independent benchmark indices
| Model Details | |||
|---|---|---|---|
| General Info | |||
| Provider | MiniMax | Anthropic | Anthropic |
| Release Date | June 1, 2026 | September 22, 2026 | September 28, 2026 |
| Knowledge Cutoff | - | - | Jun 2026 |
| Context & Limits | |||
| Context Window | 1.0M Best Context Window | 1M | 1M |
| Pricing (per 1M tokens) | |||
| Input Pricing | $0.30 Best Input Pricing | $4 | $2 |
| Output Pricing | $1.20 Best Output Pricing | $20 | $10 |
| Modalities | |||
| Inputs | textimagevideo | textimagefile | textimagefile |
| Outputs | text | text | text |
| Benchmarks (0-100) | |||
| Intelligence Index | 29.2 | 57.6 Best Intelligence Index | 56.0 |
| Coding Index | 58.6 | - | - |
| Agentic Index | 29.5 | - | - |
Humanity's Last Exam
Extremely difficult logical reasoning and knowledge.
Long Context Reasoning
Logical reasoning over long context windows.
SciCode Benchmark
Scientific coding and mathematical modeling.
Independent evaluation data provided by Artificial Analysis. To view the latest benchmarks and full details, visit their official site.
Explore more from MiniMax
Other models by MiniMax
Top AI Models
Leading alternatives by intelligence score
Claude Opus 5.5
Muse Spark 1.3
GPT-6 Astra
