MiniMax M3

ServerlessDedicated
Hugging Face IDMiniMaxAI/MiniMax-M3View on Hugging Face ↗

MiniMax M3 is MiniMax's natively multimodal model for long-horizon agentic work, including coding. It has about 428B parameters, about 23B active per token, and a 1M-token context window.

Ways to run MiniMax M3

Dedicated endpoint

Get a quote →

Your own GPUs, tuned for your traffic. We profile your workload, tune the deployment and put throughput and latency in your contract.

Serverless

API docs →

Pay per token on shared GPUs, with no capacity to plan or reserve.

Inside your network

Security →

Run a dedicated endpoint inside your own network, including air-gapped environments.

Pricing

Dedicated endpoint
Per GPU-hour
We size the GPUs to your traffic.
Serverless
$0.075 / $0.30
Per 1M tokens (input/output)

Use it with the agents we set up for you

Point any open-source agent at your endpoint, or pick from our catalog and we set it up as part of your contract.

Cline CodingOpenCode CodingAider Coding
GPT Researcher ResearchRAGFlow DocumentsNanobrowser Browser

MiniMax M3 FAQ

How much does MiniMax M3 cost on Saturn Cloud?

Serverless costs $0.075 per 1M input tokens and $0.30 per 1M output tokens. A dedicated endpoint is priced per GPU-hour for the GPUs it runs on, and we size the GPUs to your traffic.

What is MiniMax M3's context length?

1M tokens. MiniMax reports 9x faster prefill and 15x faster decode than MiniMax M2 at 1M context.

How does thinking work in MiniMax M3?

Set the thinking parameter to enabled, adaptive or disabled. With adaptive, the model decides when to think.

What is MiniMax M3 good for?

Long-horizon agentic work, especially coding. MiniMax also lists multimodal understanding across text, images and video, and long-context processing.

What sampling settings does MiniMax recommend?

Temperature 1.0 and top_p 0.95.