AA Omniscience - Software Engineering (SWE) - TypeScript: leaderboard
TypeScript-language slice of AA Omniscience SWE, testing factual knowledge of TypeScript syntax, type-system behavior, tooling, and web 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) | 77.78 | #30 (Kimi K3) |
| 2 | Kimi K3 (Low) | 75.28 | #30 (Kimi K3) |
| 3 | Claude Opus 4.5 (Thinking) | 71.11 | #83 (Claude Opus 4.5) |
| 4 | Kimi K2.7 Code | 66.67 | #89 |
| 5 | Claude Opus 4.5 (Non-reasoning) | 64.44 | #83 (Claude Opus 4.5) |
| 6 | Gemini 3 Pro (Preview) (High) | 64.44 | #66 (Gemini 3 Pro (Preview)) |
| 7 | Kimi K2.5 (Thinking) | 60 | #144 (Kimi K2.5) |
| 8 | Inkling (xHigh) | 60 | #126 (Inkling) |
| 9 | DeepSeek V4 Pro (Reasoning, Max Effort) | 57.78 | #102 (DeepSeek V4 Pro) |
| 10 | Kimi K2.6 | 54.44 | #104 |
| 11 | DeepSeek V4 Flash (Reasoning, Max Effort) | 54.44 | #137 (DeepSeek V4 Flash) |
| 12 | Hy3 | 53.33 | #93 |
| 13 | Claude Sonnet 4.5 (Thinking) | 52.22 | #142 (Claude Sonnet 4.5) |
| 14 | Qwen3.8 2.4T A95B | 52.22 | #55 |
| 15 | DeepSeek V4 Flash (Reasoning, High Effort) | 50 | #137 (DeepSeek V4 Flash) |
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-typescript · How It Works · Data refreshed daily, snapshot 2026-10-20.