Best AI models for science
Ranked by GPQA Diamond: very hard multiple-choice questions in biology, physics and chemistry written by PhD experts, where random guessing scores 25%. Tested independently by Epoch AI.
Our picks
The ranking
Compare the top two| # | Model | GPQA Diamondi | Overall indexi | Inputi | Outputi | Contexti |
|---|---|---|---|---|---|---|
| 1 | GPT-6 AstraOpenAI | 95.8% | 166.6 | $10 | $50 | 1.1M |
| 2 | Claude Sonnet 5.5Anthropic | 95.6% | — | $2 | $10 | 1M |
| 3 | Gemini 3.8 FlashGoogle DeepMind | 95.4% | 157.1 | $0.75 | $3.75 | 1M |
| 4 | Gemini 3.7 FlashGoogle DeepMind | 94.8% | 157.7 | $0.75 | $3.75 | 1M |
| 5 | GPT-5.4 ProOpenAI | 94.6% | 159.1 | $30 | $180 | 1.1M |
| 6 | Gemini 3.1 Pro PreviewGoogle DeepMind | 94.4% | 154.9 | $2 | $12 | 1M |
| 7 | Gemini 3.6 FlashGoogle DeepMind | 94.1% | 154.4 | $0.75 | $3.75 | 1M |
| 8 | Grok 4.6xAI | 94.0% | 156.5 | $2 | $6 | 500K |
| 9 | Claude Opus 5Anthropic | 93.9% | 162.7 | $5 | $25 | 1M |
| 10 | GPT-5.6 SolOpenAI | 93.5% | 162.0 | $2 | $10 | 1.1M |
per 1M tokens · USD
Quick answers
- Which AI model is best for science questions?
- GPT-6 Astra from OpenAI leads this ranking (GPQA Diamond: 95.8%). It costs $10 per million input tokens and $50 per million output tokens.
- Is there a cheaper alternative?
- Gemini 3.8 Flash from Google DeepMind (GPQA Diamond: 95.4%) costs $0.75 / $3.75 per million tokens, about 13× less than GPT-6 Astra.
How we rank
Scores come from tests run independently by Epoch AI (CC BY 4.0), not from the companies’ own announcements. Prices come from OpenRouter’s public API. Both are refreshed automatically several times a day, so this ranking follows new releases and price changes.
Other rankings
CodingMathCheapFreeOpen sourceLong contextImages & PDFsRankings →