TinyGrafixBench - Line Chart: leaderboard

TinyGrafixBench mIoU (%) on line chart edits (drawing segments, normalizing, filtering and shading a series), scored pixel by pixel against the single correct image; higher is better (PaintBench paper, NYU, 2026).

Metric: mIoU (%). Source: arxiv.org. Saturation forecast: Rough model projection: around 2026. 11 models tracked.

Top models

#ModelScoreOverall rank
1GPT-Image-216
2Nano-Banana-211.5
3BAGEL-7B-MoT4.9#5515
4FLUX.2-klein-9B4.9
5Nano Banana4.7
6FLUX.1-Kontext-dev4.5
7Qwen-Image-Edit-25114.4
8FLUX.2-dev3.8
9LongCat-Image-Edit3.2
10InstructPix2Pix0.1

Interactive version: theaggregate.ai/benchmark?slug=tinygrafixbench-line-chart · How It Works · Data refreshed daily, snapshot 2026-11-12.