Scale AI - VISTA: leaderboard
Vision-language understanding benchmark for multimodal models. Part of Scale's SEAL evaluation platform.
Metric: Score. Source: scale.com. Source: scale.com. Source: benchmarklist.com. 63 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 2.5 Pro (Preview 03-25) | 54.65 | #123 |
| 2 | Gemini 2.5 Pro (Preview 06-05) | 54.63 | #106 |
| 3 | O4 Mini (High) | 51.79 | #180 (O4 Mini) |
| 4 | O4 Mini (Medium) | 51.66 | #180 (O4 Mini) |
| 5 | Gemini 3 Pro (Preview) | 51.49 | #66 |
| 6 | GPT-5.4 (2026-03-05) (xHigh) | 50.89 | #59 (GPT-5.4 (2026-03-05)) |
| 7 | Gemini 2.5 Pro (Preview 05-06) | 50.78 | #147 |
| 8 | GPT-5 Mini (2025-08-07) | 50.39 | #172 |
| 9 | O3 (High) | 50.07 | #124 (O3) |
| 10 | GPT-5 | 49.69 | #98 |
| 11 | O3 (Medium) | 49.59 | #124 (O3) |
| 12 | Gemini 2.5 Flash (Preview 05-20) | 49.15 | #216 |
| 13 | Claude Sonnet 4.5 (Thinking) | 48.75 | #142 (Claude Sonnet 4.5) |
| 14 | Claude Opus 4.1 (20250805) (Thinking) | 48.44 | #132 (Claude Opus 4.1 (20250805)) |
| 15 | O1 Pro | 47.32 | #228 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=scale-ai-vista · How It Works · Data refreshed daily, snapshot 2026-10-20.