- Intelligence
- 29.2
- Rank 133 of 744 measured models
- Coding
- 58.6
- Rank 93 of 298 measured models
- Output speed
- 99.4 t/s
- Rank 169 of 403 measured models
- Price/1M
- $0.525
- Rank 283 of 659 measured models
- Context window
- 1M
- Rank 84 of 1199 measured models
- Cost per task
- $0.508
- Rank 105 of 193 measured models
MiniMax-M3 scores 29.2 on the Artificial Analysis Intelligence Index, ranking 133rd of 744 measured models. The closest models on this index are Gemini 3.1 Pro Preview (29.7), Qwen3.7 Max (29.5), and Claude Opus 4.5 (batch) (29.1). At $0.525 per million tokens (3:1 input/output blended price), it is cheaper than Gemini 3.1 Pro Preview ($4.50) and Qwen3.7 Max ($3.75). With an output speed of 99.4 t/s, it is slower than Gemini 3.1 Pro Preview (131 t/s) and Qwen3.7 Max (191 t/s).
Additional benchmarks15
Design Arena · 3d50.2%
Design Arena · Agenticgamedev44.6%
Design Arena · Androidnative44.3%
Design Arena · Asciiart46.8%
Design Arena · Codecategories50.9%
Design Arena · Dataviz51.3%
Design Arena · Fullstack47.3%
Design Arena · Gamedev46.0%
Design Arena · Htmlslides46.6%
Design Arena · Mobileapps45.9%
Design Arena · Python Pptxslides46.8%
Design Arena · Svg45.2%
Design Arena · Uicomponent50.9%
Design Arena · Webapps47.1%
Design Arena · Website51.9%
Models at a similar level
The closest models on the Intelligence Index, one configuration per family.
- Gemini 3.1 Pro Preview
Google · $4.50 · 131 t/s
29.7+0.5$4.50131 t/sCompare - Qwen3.7 Max
Alibaba · $3.75 · 191 t/s
29.5+0.3$3.75191 t/sCompare - Claude Opus 4.5 (batch)
Anthropic · — · —
29.1−0.1——Compare - MiMo-V2-Pro
Xiaomi · — · —
28.6−0.6——Compare
