AIME leaderboard
- models measured
- 54
- lead over #2
- +1.2 pts
- median
- 86.3%
Leader on AIME · read October 21
On October 21, 2026, 54 of our catalog's models had a AIME result, and Google Gemini 3.1 Pro Preview led with 98.1%.
AIME is the American Invitational Mathematics Examination, a competition for strong high school students. Each answer is a whole number, so grading is exact. The number is the share of problems solved, as run by Vals AI.
Role in the index
Limit: The questions and answers are public, so a model may have seen them in training. Vals AI notes that models tend to score better on the older 2024 set than on the 2025 set.
We collect this benchmark and show it as supporting evidence. It is not part of the Vibe Coding Index, because it measures something other than coding and the index is about coding. Read the methodology
| # | Model | Result | Latency, as printed | Evidence | Coding index | Output $/Mtok |
|---|---|---|---|---|---|---|
| 1 | Gemini 3.1 Pro PreviewGoogle | 98.1% | 1m 45s | 35.3 | $12.00 | |
| 2 | GPT-5.2OpenAI | 96.9% | 2m 00s | 58.9 | $14.00 | |
| 3 | GPT-5.4OpenAI | 96.7% | 2m 01s | 37.7 | $15.00 | |
| 4 | Grok 4.20xAI | 96.5% | 1m 47s | Not benchmarked | $2.50 | |
| 5 | Claude Opus 4.7Anthropic | 96.3% | 3m 02s | 51.1 | $25.00 | |
| 6 | Claude Opus 4.6Anthropic | 95.6% | 4m 56s | 66.9 | $25.00 | |
| 7 | Gemini 3 Flash PreviewGoogle | 95.6% | 1m 24s | 75.8 | $3.00 | |
| 8 | Kimi K2.5Moonshot AI | 95.6% | 5m 23s | 58.5 | $2.50 | |
| 9 | GPT-5.4 MiniOpenAI | 95.6% | 5m 00s | 28.2 | $4.50 | |
| 10 | Claude Opus 4.5Anthropic | 95.4% | 3m 32s | 66.8 | $25.00 |
Only models in our catalog are listed, and each appears once with its best published run. 54 rows were read from the source on October 21, 2026, and rows that match no catalog model are left out. Every other column is a figure the board publishes for the same run.
Cheapest good model on AIME · Best open-weight model on AIME