# FrontierMath (Tier 1–3) — AI model rankings

Very hard, research-level math problems. Tiers 1–3 are the (still extremely difficult) lower tiers. Higher is better.

6 tracked models have a published FrontierMath (Tier 1–3) score. Higher is better. Scores are as published at each model's release.

## Ranking

| Rank | Model | Developer | Score | Released |
| --- | --- | --- | --- | --- |
| 1 | GPT-5.5-Pro | OpenAI | 52.4% | Apr 23 2026 |
| 2 | GPT-5.5 | OpenAI | 51.7% | Apr 23 2026 |
| 3 | GPT-5.4-Pro | OpenAI | 50% | Mar 5 2026 |
| 4 | GPT-5.4 | OpenAI | 47.6% | Mar 5 2026 |
| 5 | Claude Opus 4.7 | Anthropic | 43.8% | Apr 16 2026 |
| 6 | Gemini 3.1 Pro | Google | 36.9% | Feb 19 2026 |


---

Canonical page: https://aireleasetracker.com/benchmark/frontiermath-t13
Full dataset: https://aireleasetracker.com/llms-full.txt · JSON: https://aireleasetracker.com/models.json
Source: AI Release Tracker (https://aireleasetracker.com). Benchmark scores are the figures published by the releasing lab at launch.