The AI Loadout
←Back to models
Checked Active: 2026-10-25
MiniMax

MiniMax M3.1 Flash

Low-cost coding agents, computer use and long multi-step tool work, with text, image and video input.

Evaluation Analysis: Released Oct 19, 2026. MiniMax says it was the anonymous Space Bunny model that processed more than 200 trillion tokens before launch, and that agent tasks cost about 70% less than on the previous generation with online inference nearly 3x faster. MiniMax reports 73.7% on DeepSWE v1.1, 61.7 on OSWorld V2 and 82% on CyberGym. Thinking is always on, with effort from low to max (max is the default). Cached input $0.015. Anthropic- and OpenAI-compatible endpoints (model MiniMax-M3.1-Flash). Also on Ollama Cloud as minimax-m3.1-flash:cloud at the same price, free there until Oct 22, 2026 (UTC). Open weights are on Hugging Face (MiniMaxAI/MiniMax-M3.1-Flash, about 428B total and 23B active parameters) under the MiniMax Community License: free for non-commercial use; commercial use requires a "Built with MiniMax M3.1-Flash" credit and a one-time notice to MiniMax, and products over $20M a year in revenue need written permission.
Context Window1M tokens
Input Price / 1M$0.15
Output Price / 1M$0.60
Release Date2026-10-19

Specs and prices checked 2026-10-25 against the official page.

Used in these setups

Related skills and personas

Agent skills and bot personas that pair well with MiniMax M3.1 Flash.

Browse full directories:All Skills (>35)·All Personas (>15)

Want tested skills and personas installed in one command? See the AI Loadout packs, from $9.