AA Omniscience - Software Engineering (SWE) - R: leaderboard
R-language slice of AA Omniscience SWE, testing factual knowledge of R syntax, packages, tooling, and data-science engineering practice.
Metric: Accuracy (%). Source: artificialanalysis.ai. Saturation forecast: Around January 2027. 294 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Kimi K3 (Max) | 64 | #30 (Kimi K3) |
| 2 | Claude Opus 4.5 (Thinking) | 64 | #83 (Claude Opus 4.5) |
| 3 | Claude Opus 4.5 (Non-reasoning) | 60 | #83 (Claude Opus 4.5) |
| 4 | Kimi K3 (Low) | 60 | #30 (Kimi K3) |
| 5 | Claude Sonnet 4.5 (Thinking) | 56 | #142 (Claude Sonnet 4.5) |
| 6 | Gemini 3 Pro (Preview) (High) | 50 | #66 (Gemini 3 Pro (Preview)) |
| 7 | Kimi K2.7 Code | 44 | #89 |
| 8 | Gemini 2.5 Pro | 42 | #152 |
| 9 | O3 | 42 | #124 |
| 10 | GPT-5.1 (High) | 42 | #135 (GPT-5.1) |
| 11 | Inkling (xHigh) | 42 | #126 (Inkling) |
| 12 | Grok 4 | 40 | #181 |
| 13 | Kimi K2.5 (Thinking) | 40 | #144 (Kimi K2.5) |
| 14 | MiMo-V2.5-Pro (Non-reasoning) | 40 | #158 (MiMo-V2.5-Pro) |
| 15 | Claude Sonnet 4.5 (Non-reasoning) | 38.78 | #142 (Claude Sonnet 4.5) |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=aa-omniscience-software-engineering-swe-r · How It Works · Data refreshed daily, snapshot 2026-10-20.