GIM: leaderboard
Grounded Integration Measure from Meta FAIR: 820 multimodal and text-grounded problems requiring integrated reasoning across quantitative, spatial, language, world-knowledge, and document tasks.
Metric: IRT ability (theta). Source: huggingface.co. Source: arxiv.org. Saturation forecast: Around April 2027. 46 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.4 Pro (xHigh) | 2.16 | #32 (GPT-5.4 Pro) |
| 2 | Gemini 3.1 Pro (Preview) (High) | 2.04 | #58 (Gemini 3.1 Pro (Preview)) |
| 3 | Gemini 3.1 Pro (Preview) (Medium) | 1.85 | #58 (Gemini 3.1 Pro (Preview)) |
| 4 | GPT-5.4 (xHigh) | 1.64 | #79 (GPT-5.4) |
| 5 | GPT-5.4 (High) | 1.59 | #79 (GPT-5.4) |
| 6 | Claude Opus 4.7 (High) | 1.53 | #49 (Claude Opus 4.7) |
| 7 | Claude Opus 4.6 (High) | 1.48 | #63 (Claude Opus 4.6) |
| 8 | Gemini 3.1 Pro (Preview) (Low) | 1.45 | #58 (Gemini 3.1 Pro (Preview)) |
| 9 | GPT-5.4 (Medium) | 1.39 | #79 (GPT-5.4) |
| 10 | Claude Sonnet 4.6 (High) | 1.12 | #90 (Claude Sonnet 4.6) |
| 11 | Claude Opus 4.6 (Medium) | 1.09 | #63 (Claude Opus 4.6) |
| 12 | GPT-5 (High) | 0.9 | #98 (GPT-5) |
| 13 | Claude Sonnet 4.6 (Medium) | 0.84 | #90 (Claude Sonnet 4.6) |
| 14 | GPT-5.4 (Low) | 0.82 | #79 (GPT-5.4) |
| 15 | GPT-5 (Medium) | 0.59 | #98 (GPT-5) |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=gim · How It Works · Data refreshed daily, snapshot 2026-10-20.