Benchmark Almanac

SuperCLUE Arena Image Editing

ArtArk 1,340 tops its leaderboard GPT-Image-2.5 Sunburst 1,308 leads the models we track
Not in model cards

Models edit images in response to instructions. The arena ranks systems from pairwise comparisons.

The arena reports median ratings and uncertainty intervals.

Made by
SuperCLUE
Kind
Image generation and editing
Level
University
Scored
Ranked by matches
Answers
Open-ended
Tasks
All public
Measures
Preference
Language
Chinese
In and out
Text, Images → Images
Interaction
Single turn
Items from
Live play or votes
Refresh
Live
Run by
A host
Scores refreshed
November 20th, 2026
19
Entries on its own leaderboard, 15 models we track

ArtArk tops the leaderboard with 1,340 GPT-Image-2.5 Sunburst leads with 1,308

Best score per entry from the benchmark's own leaderboard, read on November 20th, 2026. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders. The best model we track is GPT-Image-2.5 Sunburst, at 1,308. This view shows only models we track, each at its best.

Bars start at 600.

  1. 1 ArtArk 1,340
  2. 2 GPT-Image-2.5 Sunburst 1,308
  3. 3 GPT-Image-2.5 Flare 1,186
  4. 4 Gemini-3.1-Flash-Image-Preview(Nano Banana 2) 1,180
  5. 5 GPT-Image-2 1,151
  6. 6 Seedream 5.0 Pro 1,145
  7. 7 Gemini 3 Pro Image Preview (nano-banana-pro) 1,128
  8. 8 GPT-Image-1.5 1,047
  9. 9 Qwen-Image-3.0-Pro 1,046
  10. 10 Seedream 4.5 965
  11. 11 Seedream 5.0 Lite 944
  12. 12 Qwen-Image-2.0-Pro 923
  13. 13 SenseNova U1.5 Lite 906
  14. 14 Wan 2.6(2025-12-16) 876
  15. 15 Gemini 2.5 Flash Image Preview (nano-banana) 815
  16. 16 GPT-Image-1 812
  17. 17 Seedream 4.0 805
  18. 18 Qwen-Image-Edit 777
  19. 19 SeedEdit 3.0 647

Source: Official SuperCLUE Arena Image Editing leaderboard

#EntryScore ±SettingDateSource
1 ArtArk 1,340 ±38.6 Official leaderboard
2 GPT-Image-2.5 Sunburst 1,308 1268.6 to 1353.1 Official leaderboard
3 GPT-Image-2.5 Flare 1,186 ±42.4 Official leaderboard
4 Gemini-3.1-Flash-Image-Preview(Nano Banana 2) 1,180 1162.7 to 1196.3 Official leaderboard
5 GPT-Image-2 1,151 1136.2 to 1165.4 Official leaderboard
6 Seedream 5.0 Pro 1,145 1128 to 1162 Official leaderboard
7 Gemini 3 Pro Image Preview (nano-banana-pro) 1,128 1110.9 to 1144.2 Official leaderboard
8 GPT-Image-1.5 1,047 1032.8 to 1061.9 Official leaderboard
9 Qwen-Image-3.0-Pro 1,046 1017.8 to 1071.9 Official leaderboard
10 Seedream 4.5 965 948.2 to 980.7 Official leaderboard
11 Seedream 5.0 Lite 944 927 to 959.3 Official leaderboard
12 Qwen-Image-2.0-Pro 923 ±15.9 Official leaderboard
13 SenseNova U1.5 Lite 906 ±28.4 Official leaderboard
14 Wan 2.6(2025-12-16) 876 862.2 to 891.2 Official leaderboard
15 Gemini 2.5 Flash Image Preview (nano-banana) 815 802.3 to 828.4 Official leaderboard
16 GPT-Image-1 812 797.6 to 824.4 Official leaderboard
17 Seedream 4.0 805 ±13.2 Official leaderboard
18 Qwen-Image-Edit 777 ±14 Official leaderboard
19 SeedEdit 3.0 647 ±13.4 Official leaderboard
#ModelScore ±SettingDateSource
1 GPT-Image-2.5 Sunburst 1,308 1268.6 to 1353.1 Official leaderboard
2 GPT-Image-2.5 Flare 1,186 ±42.4 Official leaderboard
3 GPT-Image-2 1,151 1136.2 to 1165.4 Official leaderboard
4 Seedream 5.0 Pro 1,145 1128 to 1162 Official leaderboard
5 Gemini 3 Pro Image Preview (nano-banana-pro) 1,128 1110.9 to 1144.2 Official leaderboard
6 GPT-Image-1.5 1,047 1032.8 to 1061.9 Official leaderboard
7 Qwen-Image-3.0-Pro 1,046 1017.8 to 1071.9 Official leaderboard
8 Seedream 4.5 965 948.2 to 980.7 Official leaderboard
9 Seedream 5.0 Lite 944 927 to 959.3 Official leaderboard
10 Qwen-Image-2.0-Pro 923 ±15.9 Official leaderboard
11 Gemini 2.5 Flash Image Preview (nano-banana) 815 802.3 to 828.4 Official leaderboard
12 GPT-Image-1 812 797.6 to 824.4 Official leaderboard
13 Seedream 4.0 805 ±13.2 Official leaderboard
14 Qwen-Image-Edit 777 ±14 Official leaderboard
15 SeedEdit 3.0 647 ±13.4 Official leaderboard

Scores over time

All the scores we have, one dot each, so a model tested at several settings appears more than once.

No dated values yet.

No lab model card reports it yet

We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.

More in Vision

All 636 in Vision
  • MMMU In use 11.5K college-level multimodal questions across 30 subjects. Multiple-choice and open accuracy. Qwen 3.6 Plus 86%
  • MMMU-Pro In use Harder MMMU: college-level multimodal questions with more options and vision-only inputs to remove text shortcuts. GPT-5.6 Sol 84.6%
  • CharXiv In use Chart understanding on real arXiv figures. Reasoning questions require synthesizing information from complex charts. Claude Fable 5 93.5%
  • MathVista In use Math reasoning in visual contexts (charts, geometry, puzzles) from 31 datasets. Answer accuracy. Seed2.1 Pro 90.7%
  • MATH-Vision (MathVision) In use 3,040 competition math problems with visual context. Answer accuracy. Qwen 3.8 Max 97.7%
  • MMStar In use Models answer 1,500 human-curated image questions that cannot be solved without the image, across 6 core capabilities. Scored by accuracy. Qwen 3.8 Max 85.9%

Data audit

All 8 checks pass. Every benchmark

Catalog record 2 pass
  • Pass Task access All tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 2 pass 3 notes
  • Pass Independent runs Scores from the official leaderboard (19 entries).
  • Pass Models scored 15 models we track, plus 4 other entries on its leaderboard.
  • Note Newest score No scores yet.
  • Note Frontier models None of its models are on the capabilities index.
  • Note Headroom Scored as a rating, with no fixed ceiling.
Leaderboard import 3 pass
  • Pass Importer We read the board on superclueai.com: 15 rows for models we track and 4 other entries at the last read.
  • Pass Last read November 20th, 2026.
  • Pass Read worked The last read worked.
Consistency 1 pass 1 note
  • Note Lab claims No model has both a lab number and an independent run.
  • Pass Known issues No published audit found problems.

How we check