RoadmapBench - UI and Rendering: leaderboard

RoadmapBench (2026) tests coding agents on 115 long-horizon tasks built from real open-source version upgrades in 17 repositories and five languages: from the earlier release and a roadmap of subtasks, the agent must implement the next release, with each subtask checked by its own tests. This board is the percentage of the 14 UI and rendering tasks the agent fully resolves, every subtask test suite passing; higher is better.

Metric: Resolved rate (%) on the 14 UI and rendering tasks (every subtask test suite passes), 115 long-horizon tasks implementing a real release upgrade from a roadmap of subtasks, one trial per model, 2-hour wall-clock budget, extended thinking on; each row names its agent scaffold (OpenHands, the primary scaffold, or Terminus 2, the ablation); higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 13 models tracked. Updated 7 Oct 2026.

Top models

#ModelSettingScoreOverall rank
1Claude Opus 4.7xHigh64.3#85
2Claude Opus 4.6Thinking57.1#113
3GPT-5.4High50#135
4Kimi K2.535.7#266
5Gemini 3.1 Pro (Preview)High35.7#99
6DeepSeek V4 ProHigh35.7#196
7Kimi K2.628.6#192
8Qwen 3.5 397B A17B21.4#264
9Qwen 3.6 Plus21.4#238
10GLM-5.121.4#198
11MiMo-V2.5-Pro21.4#284
12MiniMax-M2.714.3#539
13Seed 2.0 ProHigh14.3#222

No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=roadmapbench-ui-and-rendering · How It Works · Data refreshed daily, snapshot 2026-11-21.