- Intelligence
- 22.8
- Rank 208 of 747 measured models
- Output speed
- 66.9 t/s
- Rank 277 of 406 measured models
- Price/1M
- $0.525
- Rank 286 of 664 measured models
- Context window
- 205K
- Rank 615 of 1207 measured models
MiniMax-M2.5 scores 22.8 on the Artificial Analysis Intelligence Index, ranking 208th of 747 measured models. The closest models on this index are Nemotron 3 Ultra 550B A55B (22.9), Qwen3.5 27B (22.9), and MiniMax-M2.7 (22.8). At $0.525 per million tokens (3:1 input/output blended price), it is cheaper than Nemotron 3 Ultra 550B A55B ($1.08) and Qwen3.5 27B ($0.825), priced close to MiniMax-M2.7 ($0.525). With an output speed of 66.9 t/s, it is faster than MiniMax-M2.7 (58.9 t/s), slower than Nemotron 3 Ultra 550B A55B (156 t/s) and Qwen3.5 27B (77.3 t/s).
Additional benchmarks7
Design Arena · 3d57.6%
Design Arena · Codecategories56.8%
Design Arena · Dataviz51.2%
Design Arena · Gamedev55.5%
Design Arena · Svg54.5%
Design Arena · Uicomponent53.4%
Design Arena · Website57.5%
Models at a similar level
The closest models on the Intelligence Index, one configuration per family.
- Nemotron 3 Ultra 550B A55B
NVIDIA · $1.08 · 156 t/s
22.9+0.1$1.08156 t/sCompare - Qwen3.5 27B
Alibaba · $0.825 · 77.3 t/s
22.9+0.1$0.82577.3 t/sCompare - MiniMax-M2.7
MiniMax · $0.525 · 58.9 t/s
22.8$0.52558.9 t/sCompare - GPT-5.5 Instant (May 2026)
OpenAI · $11.25 · —
22.7−0.1$11.25—Compare
