OmniGAIA - Hard: leaderboard
The "Hard" difficulty tier of OmniGAIA (accuracy on its hard questions). Omni-modal general AI assistant benchmark for long-horizon agents using vision, audio, language, web search, browser, and code-execution tasks.
Metric: Hard-Tier Accuracy (%). Source: huggingface.co. Saturation forecast: Around February 2027. 16 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3 Pro | 38.5 | #135 |
| 2 | Gemini 3 Flash | 37.2 | #159 |
| 3 | Gemini 2.5 Pro | 21.8 | #262 |
| 4 | Gemini 2.5 Flash Lite | 7.7 | #958 |
| 5 | Qwen2.5-Omni-7B | 1.3 | #1797 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=omnigaia-hard · How It Works · Data refreshed daily, snapshot 2026-11-12.