PhyEditBench: leaderboard

Metric: Overall judge score (1-10 scale) on the 238 real-world physical-process instances: 0.2 consistency + 0.3 instruction following + 0.4 physical plausibility + 0.1 image quality, averaged over five run types (three single-step edits between consecutive states, three steps applied jointly, one global edit); GPT-4o judge scores each output 1-10 on consistency, instruction following, physical plausibility and image quality against the instruction, physical explanation, invariants and reference frames; higher is better. Source: arxiv.org. Saturation forecast: Rough model projection: around 2026. 12 models tracked.

Top models

#ModelScore
1ChronoEdit-14B8.51
2Seedream4.08.47
3GPT-Image-1.58.23
4UniWorld-V27.08
5Step1X-Edit-v1p16.79
6Qwen-Image-Edit6.71
7gemini-2.5-flash-image6.51
8BAGEL-7B-MoT (Thinking)6.43
9BAGEL-7B-MoT6.09
10InstructPix2Pix5.61

Interactive version: theaggregate.ai/benchmark?slug=phyeditbench · How It Works · Data refreshed daily, snapshot 2026-09-29.