GAIA — overall score (all levels combined) — GAIA leaderboard overall aggregate, same source as level1/2/3.
| # | Model | Vendor | Best score | Runs | Last seen |
|---|---|---|---|---|---|
| 1 | GPT-5.2 | OpenAI | 77.4 | 1 | 2026-05-10 |
| 2 | DeepSeek-V3.2 | DeepSeek | 76.7 | 1 | 2026-05-10 |
| 3 | Gemini 3 Flash Preview | 68.4 | 1 | 2026-05-11 | |
| 4 | GPT-5 | OpenAI | 66.1 | 1 | 2026-05-10 |
| 5 | GPT-4o | OpenAI | 28.9 | 1 | 2026-05-10 |
| 6 | Claude Opus 4.7 | Anthropic | 28 | 1 | 2026-06-10 |
| 7 | GPT-5 nano | OpenAI | 23.3 | 1 | 2026-05-10 |
| 8 | GPT-4 Turbo | OpenAI | 11.3 | 1 | 2026-05-11 |
| 9 | Qwen3 4B | Alibaba | 10.7 | 1 | 2026-05-19 |
| 10 | Gemma 4 26B A4B IT | 8.3 | 1 | 2026-05-16 | |
| 11 | GPT-4o mini | OpenAI | 4.7 | 1 | 2026-05-10 |
| 12 | GPT4 | OpenAI | 1 | 1 | 2026-05-10 |
| 13 | Qwen2.5 7B | Alibaba | 0.7 | 1 | 2026-05-10 |
Best tracked score per model (default configuration; source-attributed and verification-tiered). Open a model for its full benchmark surface, provenance and pricing.