TRACK - Code Generation (Facts in Context, Thinking): leaderboard
Metric: Holistic pass (%; BigCodeBench-derived coding problems that require specific library API versions, 500 items; the final answer is correct and the reasoning entails every required fact; answered with the missing facts appended to the prompt and longer thinking enabled). Source: arxiv.org. Saturation forecast: Estimated already saturated. 4 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | O4 Mini (2025-04-16) | 40.6 |
| 2 | Qwen 3 8B (Thinking) | 26.5 |
| 3 | Qwen 3 4B (Reasoning) | 22.8 |
| 4 | Qwen 3 1.7B (Thinking) | 15.5 |
Interactive version: theaggregate.ai/benchmark?slug=track-code-generation-facts-in-context-thinking · How It Works · Data refreshed daily, snapshot 2026-09-26.