MiniMax: MiniMax M3Active

minimax/minimax-m3

MiniMax M3 is the first open-weights flagship that brings coding, agentic reasoning, million-token context, and native multimodality together in one model. It handles autonomous task decomposition, tool use, and multi-step reasoning with ease — and writes code that's meant to ship, not code that just runs. Built on MiniMax's proprietary Sparse Attention architecture, M3 natively supports million-token context windows for long-horizon agents, large codebases, and long-video understanding. Its multimodality isn't bolted on — it's trained in from day one, with text and vision deeply aligned at the semantic level.

By:MiniMaxInput Type:Output Type:Publish time:2026-06-01

Providers

Route requests across multiple providers. Copy a provider slug to set your preference.

ProviderCache Hit RateContext

MiniMax

minimax

98.6%$0.3-0.6/ M tokens$1.2-2.4/ M tokensRead:0.06-0.12/ M tokensWrite:-/ M tokens1M1.03s9.34tps

Uptime

24hours

Direct request success rate on AI Gateway and per-provider.

Oct 26, 20:14

MiniMax100.00%

Throughput

24hours

P50 throughput on live AI Gateway traffic, in tokens per second (TPS).

Latency

24hours

P50 time to first token (TTFT) on live AI Gateway traffic, in milliseconds.

Oct 26, 19:00

MiniMax1.03s

Activity

Token volume and request traffic to this model over time.

Token Consumption

Oct 25

MiniMax537.6M

Token volume broken down by the provider that served each request. Larger segments indicate providers handling more of this model's traffic.

Benchmarks

Scores on standardized evaluations. Higher percentages are better — and rank percentile shows

Metrics sourced fromArtificial Analysis

Intelligence Index29.2

Apps

Public apps that send the most traffic to this model. Good signal for what real production workloads look like — and a hint at which use cases this model is best suited for. View All

Token Ranking

Pi Coding AgentPi is an open-source, terminal-based AI coding agent created by Mario Zechner (badlogic) under the earendil-works/pi project. It runs locally in a project directory with a minimal built-in toolset (read, write, edit, bash) plus session management, and is designed to be extended with TypeScript extensions, skills, prompt templates, themes and shareable Pi packages. It works with many LLM providers and can be used interactively, in print/JSON mode, over an RPC protocol, or embedded via its SDK.

55.1M36.66%

Claude CodeClaude Code is Anthropic's agentic coding tool that runs in the terminal, understands a project's codebase, edits files, runs commands, and handles workflows such as tests, debugging and git operations through natural-language instructions. It is also available as IDE extensions (VS Code, JetBrains), a desktop app and a web surface, and is used by developers to build features, fix bugs and automate routine development tasks.

43.0K10.59%

OpenQiao Studio ChatOpenQiao Studio Chat is the built-in web chat workspace of the OpenQiao AI model aggregation platform, operated by OpenQiao Singapore. It lets users converse with many different large language models in a browser GUI, with features such as multi-model side-by-side comparison, multimodal file input, Artifacts generation, web search, image generation/editing and voice input/output.

22.8K159.45%

OpenCodeOpenCode is an open-source, MIT-licensed AI coding agent built by the team behind SST (now Anomaly), available as a terminal TUI/CLI, a desktop app, and IDE extensions. It is model-agnostic and works with many LLM providers or local models, letting developers explore codebases, plan changes, edit files, run shell commands, and use LSP and MCP integrations directly from their workflow.

19.9K2.02%

Related Models

More models from MiniMax

MiniMax: MiniMax H3 Lite

-tokens

Extremely fast generation at a low price, well suited for both image-to-video and text-to-video. It also performs quite well on simple r2va tasks, making it highly suitable for large-scale content generation scenarios.

MiniMax: MiniMax H3 Highspeed

-tokens

Low cost, strong quality, and fast generation, with excellent reference-to-video performance. Ideal for scenarios that require a balanced combination of output quality, capability, and speed.

MiniMax: MiniMax M3.1-Flash

27.09Btokens

MiniMax-M3.1-Flash is a frontier multimodal language model with a 1M-token context window, designed for agentic reasoning, tool use, coding, and long-context tasks. It supports text, image, and video inputs, and offers tunable thinking depth through effort levels to balance reasoning quality, latency, and token usage.