MM-CondChain - GUI: leaderboard
Metric: Path F1 (%), the harmonic mean of True-path accuracy (follow every condition to the final answer) and False-path accuracy (stop at the perturbed condition and pick its auxiliary answer), on 377 AITZ GUI trajectories (3,421 screenshots); zero-shot multiple choice with a boxed answer, unparseable outputs wrong; higher is better. Source: arxiv.org. Saturation forecast: Around April 2027. 27 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 3.5 397B A17B | 40.19 | #141 |
| 2 | GPT-5 | 38.06 | #91 |
| 3 | Gemini 3 Pro | 38.05 | #77 |
| 4 | Gemini 3 Flash | 35.78 | #93 |
| 5 | Qwen 3.5 122B A10B | 34.17 | #170 |
| 6 | Kimi K2.5 | 33.72 | #139 |
| 7 | Qwen 3 VL 30B A3B (Thinking) | 32.93 | #338 (Qwen 3 VL 30B A3B) |
| 8 | Qwen 3 VL 8B (Thinking) | 31.83 | |
| 9 | Qwen 3 VL 235B A22B (Thinking) | 31.23 | #228 (Qwen 3 VL 235B A22B) |
| 10 | GLM-4.6V | 27.11 | #309 |
| 11 | Qwen 3 VL 235B A22B Instruct | 27.04 | #264 |
| 12 | Qwen 3.5 35B A3B | 24.02 | #250 |
| 13 | Qwen 3 VL 8B Instruct | 20.65 | #401 |
| 14 | InternVL3-38B | 20.46 | #395 |
| 15 | GPT-4o (2024-11-20) | 20.46 | #369 |
No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #8 Claude Fable 5.1.
Interactive version: theaggregate.ai/benchmark?slug=mm-condchain-gui · How It Works · Data refreshed daily, snapshot 2026-10-11.