Supabase Evals (No Skills): leaderboard
Supabase Evals without the skills pack: the same 19 CLI build tasks solved from the agent’s own knowledge of Supabase; percent of tasks with every check passed, one row per underlying model under its vendor-native harness.
Metric: Tasks Passed (%). Source: supabase.com. Saturation forecast: Around February 2027. 6 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 5.5 (Medium) | 89.9 | #3 (Claude Opus 5.5) |
| 2 | Kimi K3 | 88.4 | #35 |
| 3 | Grok 4.7 (High) | 84.1 | #26 (Grok 4.7) |
| 4 | GPT-6.1 Sol (Medium) | 78.3 | #7 (GPT-6.1 Sol) |
| 5 | Claude Sonnet 5.5 (Medium) | 76.8 | #14 (Claude Sonnet 5.5) |
| 6 | GPT-6 Luna (Medium) | 69.6 | #61 (GPT-6 Luna) |
No result here: #6 GPT-6 Astra, #9 Gemini 4 Argon, #10 Claude Fable 5.1.
Interactive version: theaggregate.ai/benchmark?slug=supabase-evals-no-skills · How It Works · Data refreshed daily, snapshot 2026-11-05.