MiniMax

MiniMax-M3

Generally availableOpen weights1M contextReasoning

Summary

MiniMax-M3 is an open-weight language model from MiniMax, released on 1 Jun 2026. It accepts text, images and video and generates text, with a 1M-token context window and up to 524K output tokens per response. As of 18 Nov 2026, the lowest listed API price is $0.28 per million input tokens and $1.10 per million output tokens (DeepInfra), across 10 providers we track.

MiniMax multimodal model for long-context coding, perception, and agent planning

Last verified 18 Nov 2026. Every figure on this page links to its source.

Specifications

DeveloperMiniMax
Release date1 Jun 2026 [models.dev]
StatusGenerally available [models.dev]
WeightsOpen weights [Hugging Face]
Licenseminimax-community [Hugging Face]
Parameters427B [Hugging Face]
Context window1M (1,000,000 tokens) [models.dev]
Max output524K tokens [models.dev]
Knowledge cutoffJan 2025 [models.dev]
InputText, images and video [models.dev]
OutputText [models.dev]
Reasoning modeYes [models.dev]
Tool callingYes [models.dev]
Structured outputNo [models.dev]
Hugging FaceMiniMaxAI/Minimax-M3 [Hugging Face]

MiniMax-M3 API pricing by provider

USD per million tokens. Sorted by input price.

ProviderInputOutputCached inputContextSource
DeepInfra
MiniMaxAI/MiniMax-M3
$0.28$1.10$0.06524Kmodels.dev, 18 Nov 2026
MiniMax API
MiniMax-M3
$0.30$1.20$0.061Mmodels.dev, 18 Nov 2026
Fireworks AI
accounts/fireworks/models/minimax-m3
$0.30$1.20$0.06512Kmodels.dev, 18 Nov 2026
Hugging Face Inference Providers
MiniMaxAI/MiniMax-M3
$0.30$1.20–524Kmodels.dev, 18 Nov 2026
Novita AI
minimax/minimax-m3
$0.30$1.20$0.061Mmodels.dev, 18 Nov 2026
Nebius AI Studio
MiniMaxAI/MiniMax-M3
$0.30$1.20$0.061Mmodels.dev, 18 Nov 2026
Fireworks AI
accounts/fireworks/routers/minimax-latest
$0.30$1.20$0.06512Kmodels.dev, 18 Nov 2026
SiliconFlow
MiniMaxAI/MiniMax-M3
$0.30$1.20$0.061Mmodels.dev, 18 Nov 2026
Together AI
MiniMaxAI/MiniMax-M3
$0.30$1.20$0.06524Kmodels.dev, 18 Nov 2026
OpenRouter
minimax/minimax-m3
$0.30$1.20$0.06524KOpenRouter, 18 Nov 2026
Vercel AI Gateway
minimax/minimax-m3
$0.30$1.20$0.061Mmodels.dev, 18 Nov 2026

Running MiniMax-M3 locally

Estimated memory, from the parameter count and the architecture in config.json on Hugging Face. Includes about 0.8 GB of runtime overhead.

QuantizationWeightsTotal, 8K contextTotal, 32K context
Q4_K_M (4-bit)259 GB261 GB264 GB
Q8_0 (8-bit)454 GB456 GB459 GB
FP16 / BF16 (16-bit)854 GB856 GB859 GB

At 4-bit with an 8K conversation, MiniMax-M3 needs about 261 GB, more than any single consumer graphics card; it needs a unified-memory machine, a multi-GPU setup or a split between graphics card and system RAM. Check your own hardware, or see Best computer for local AI on our sister site AIPCs.

Frequently asked questions

What is the context window of MiniMax-M3?

MiniMax-M3 has a context window of 1,000,000 tokens (1M) and can generate up to 524,288 tokens in a single response, according to models.dev.

How much does the MiniMax-M3 API cost?

Through MiniMax API, MiniMax-M3 costs $0.30 per million input tokens and $1.20 per million output tokens. The lowest listed price is $0.28 input / $1.10 output through DeepInfra. Prices last verified 18 Nov 2026.

Is MiniMax-M3 open source?

MiniMax-M3 is an open-weight model: its weights are published on Hugging Face as MiniMaxAI/Minimax-M3 under the minimax-community license. Check the license terms before commercial use.

When was MiniMax-M3 released?

MiniMax released MiniMax-M3 on 1 Jun 2026, according to models.dev.

Which providers offer MiniMax-M3?

We track MiniMax-M3 on 10 providers: DeepInfra, MiniMax API, Fireworks AI, Hugging Face Inference Providers, Novita AI, Nebius AI Studio, SiliconFlow, Together AI, OpenRouter, Vercel AI Gateway.

How much VRAM does MiniMax-M3 need?

About 261 GB at 4-bit (Q4_K_M) with an 8K-token conversation, and about 456 GB at 8-bit. That is more than any single consumer graphics card holds. These are estimates from the model's published parameter count and architecture.

Similar models from other labs

Comparable on output type, weights, release date, price and context window. Prices are the lowest listed, USD per million tokens.

ModelLabReleasedContextInput $/1MOutput $/1MWeights
LongCat-2.0Meituan30 Jun 20261M$0.30$1.20Open
GLM-5.3Zhipu AI (Z.ai)14 Aug 20261M$0.07$1.40Open
Inkling SmallThinkingmachines30 Jul 20261M$0.45$1.20Open
GLM-5.2Zhipu AI (Z.ai)13 Jun 20261M$0.007$2.40Open
Nemotron 3 UltraNVIDIA4 Jun 20261M$0.50$2.20Open
MiMo-V2.6-ProXiaomi (MiMo)22 Sept 20261M$0.43$0.87Open

Change history

  • Nebius AI Studio price changed: input $0.30 → $0.30, output $1.20 → $1.20 per 1M tokens
  • Nebius AI Studio price changed: input $0.30 → $0.30, output $1.20 → $1.20 per 1M tokens
  • Max output changed from 512K to 524K
  • Structured output changed from none to no
  • Now listed on Novita AI

Sources