Best LLMs for Science — MiniMax
The best MiniMax model for science is MiniMax M3, scoring 92.9% on GPQA Diamond — graduate-level physics, chemistry and biology. MiniMax M2.7 (87.4%) and MiniMax M2.5 (84.8%) round out the top three.
- 1
minimax-m3ReasoningToolsJSON+144.4 intel · $0.300/M · 101 t/s92.9%GPQA - 2minimax-m2.7ReasoningToolsJSON38.1 intel · $0.270/M · 261 t/s87.4%GPQA
- 3minimax-m2.5ReasoningToolsJSON33.7 intel · $0.150/M · 92 t/s84.8%GPQA
- 4minimax-m2.1ReasoningToolsJSON31.4 intel · $0.300/M · 63 t/s83.0%GPQA
- 5minimax-m2ReasoningToolsJSON28.3 intel · $0.255/M · 48 t/s77.7%GPQA
How this is ranked
AI models ranked by GPQA Diamond — graduate-level physics, chemistry and biology questions written to be un-Googleable, so the score reflects reasoning from knowledge rather than retrieval. The best large language models for scientific research and hard-science work.
Frequently asked
What is the best MiniMax model for science?
The best MiniMax model for science is MiniMax M3, scoring 92.9% on GPQA Diamond — graduate-level physics, chemistry and biology. MiniMax M2.7 (87.4%) and MiniMax M2.5 (84.8%) round out the top three.
What's a good alternative to MiniMax M3?
MiniMax M2.7 (87.4%) is the closest alternative on this metric, followed by MiniMax M2.5 (84.8%). See the full ranking above for the tradeoffs.
How many MiniMax models are there?
modelgrep tracks 8 MiniMax models with live benchmarks, speed, latency and per-provider pricing, led on intelligence by MiniMax M3. 5 of them qualify for this ranking.