modelgrep
M

MiniMax: MiniMax M3

minimax/minimax-m3

7th smartest of 178ReasoningToolsJSONVision
Use via OpenRouter ↗
Intelligence
54.7
7th of 178
Design Elo
1330
3D
Speed
50
164th fastest
Latency
630ms
first token
Input price
$0.300
128th cheapest
Context
1.0M
512K max out

How it compares

Smarter than96%
of all ranked models
Faster than44%
of all ranked models
Cheaper than57%
of all ranked models

Overview

MiniMax-M3 is a multimodal foundation model from MiniMax. It supports text, image, and video inputs with text output, a 1M-token context window, and is suited for long-horizon agentic work, coding,...

Benchmarks

independent · via OpenRouter
Artificial Analysis96th percentile
Intelligence Index
54.7
Coding Index
43.4
Agentic Index
68.6
GPQA Diamond
93%
Humanity's Last Exam
37%
SciCode
45%
Tau²-Bench (agentic)
89%
Design Arena · Elo3,424 tournaments
3D
1330
codecategories
1311
Website
1306
UI Component
1302
Data Viz
1292
Game Dev
1272
svg
1256

Providers & pricing (5)

ProviderIn $/MOut $/MUptime
Minimaxfp8$0.300$1.2099.8%
Novita$0.300$1.2097.9%
Parasailfp8$0.300$1.2094.1%
AtlasCloudfp8$0.420$1.6899.8%
Morph$0.600$2.4099%

Specifications

Context window1.0M
Max output512K
Knowledge cutoff
Input modalitiestext, image, video
Output modalitiestext
Prompt caching
Cache read price$0.060/M
ModeratedNo

MiniMax M3 FAQ

How much does MiniMax M3 cost?

MiniMax M3 costs $0.300 per million input tokens and $1.20 per million output tokens via OpenRouter, making it 128th cheapest of 298 paid models.

How smart is MiniMax M3?

MiniMax M3 scores 54.7 on the Artificial Analysis Intelligence Index, ranking 7th of 178 benchmarked models, with a GPQA Diamond score of 93%.

How fast is MiniMax M3?

MiniMax M3 generates around 50 tokens per second with 630ms time-to-first-token (p50), the 164th fastest tracked model.

What is MiniMax M3's context window?

MiniMax M3 supports a 1.0M-token context window and can output up to 512K tokens. It accepts text, image, video input.

Compare head-to-head