Best LLMs for Science — SpaceXAI
The best SpaceXAI model for science is Grok 4.5, scoring 93.1% on GPQA Diamond — graduate-level physics, chemistry and biology. Grok 4.20 (91.1%) and Grok 4.3 (90.1%) round out the top three.
- 1
grok-4.5ReasoningToolsJSON+153.8 intel · $2.00/M · 52 t/s93.1%GPQA - 2grok-4.20ReasoningToolsJSON+137.0 intel · $1.25/M · 144 t/s91.1%GPQA
- 3grok-4.3ReasoningToolsJSON+137.6 intel · $1.25/M · 95 t/s90.1%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 SpaceXAI model for science?
The best SpaceXAI model for science is Grok 4.5, scoring 93.1% on GPQA Diamond — graduate-level physics, chemistry and biology. Grok 4.20 (91.1%) and Grok 4.3 (90.1%) round out the top three.
What's a good alternative to Grok 4.5?
Grok 4.20 (91.1%) is the closest alternative on this metric, followed by Grok 4.3 (90.1%). See the full ranking above for the tradeoffs.
How many SpaceXAI models are there?
modelgrep tracks 5 SpaceXAI models with live benchmarks, speed, latency and per-provider pricing, led on intelligence by Grok 4.5. 3 of them qualify for this ranking.