Benchmark Almanac

DesignArena Image Editing

GPT-Image-2.5 Sunburst 1,370
Not in model cards

Image models edit a user image following a text instruction. Users compare the two edits and vote for the better one.

One of 40 DesignArena boards. DesignArena ranks entries with Bradley-Terry ratings from user votes.

Kind
Image generation and editing
Level
Everyday
Scored
Ranked by votes
Answers
Open-ended
Tasks
All public
Last model added
November 2026
Scores refreshed
November 12th, 2026
54
Entries on its own leaderboard, 21 models we track

GPT-Image-2.5 Sunburst leads with 1,370

Best score per entry from the benchmark's own leaderboard, read on November 12th, 2026. Where a model ran with several settings or harnesses, we keep its best. Every run is in the next chart. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders.

54 entries. Scroll sideways for the rest. Bars start at 600.

  1. 1,370 GPT-Image-2.5 Sunburst
  2. 1,349 GPT-Image-2.6 Sunburst
  3. 1,341 GPT-Image-2.5 Flare
  4. 1,338 GPT-Image-2.6 Flare
  5. 1,315 GPT-Image-2
  6. 1,286 Riverflow 2.5 Pro
  7. 1,279 Krea 3
  8. 1,271 Reve 2.0
  9. 1,268 Reve 2.1
  10. 1,268 ChatGPT Image
  11. 1,254 Grok Imagine Image 2 Auto
  12. 1,251 Grok Imagine Image 2.0
  13. 1,249 Muse Image
  14. 1,247 FLUX 3 Image Edit
  15. 1,239 Seedream 5.0 Pro Edit
  16. 1,235 MAI-Image-2.6 Flash
  17. 1,228 GPT-Image-1.5
  18. 1,228 Qwen-Image-3.0-Pro
  19. 1,228 MiniMax H3-Image
  20. 1,227 Gemini 3 Pro Image Preview 2K
  21. 1,224 MAI-Image-2.6
  22. 1,220 Gemini 3 Pro Image Preview
  23. 1,218 Ideogram 4.5
  24. 1,202 MAI-Image-2.5
  25. 1,177 Grok Imagine
  26. 1,169 Gemini 3.1 Flash-Lite Image (nano-banana-2-lite)
  27. 1,169 UNI-1.1
  28. 1,167 Seedream Lite 5.0 Edit
  29. 1,155 FLUX 3 Image Klein Edit
  30. 1,146 UNI-1.1 Max
  31. 1,141 Seedream 4.5 Edit
  32. 1,126 Wan 2.7
  33. 1,120 Wan 2.7 Pro
  34. 1,111 FLUX.2 Pro Edit
  35. 1,110 DreamImage 2.0
  36. 1,102 ImagineArt 2.0 Edit
  37. 1,097 Seedream 4.0 Edit 4K
  38. 1,095 Seedream 4.0 Edit
  39. 1,081 GPT-Image-1
  40. 1,075 FLUX.2 Klein 9B Edit
  41. 1,068 Reve Edit
  42. 1,064 Qwen-Image-Edit (2511)
  43. 1,045 Reve Edit Fast
  44. 1,043 P-Image-Edit
  45. 1,032 FLUX.2 Klein 4B Edit
  46. 1,031 FLUX.2 Klein 4B Distilled Edit
  47. 1,018 FLUX Kontext Max Edit
  48. 1,011 Boogu-Image-0.1
  49. 1,002 FLUX Kontext Pro Edit
  50. 1,001 Qwen-Image-Edit
  51. 972 HiDream O1 Image Edit
  52. 961 HiDream O1 Image Dev Edit
  53. 906 Gemini 2.0 Flash Edit
  54. 787 Recraft V3 Edit

Bars start at 600.

  1. 1 GPT-Image-2.5 Sunburst 1,370
  2. 2 GPT-Image-2.6 Sunburst quality: max, complexity: high 1,349
  3. 3 GPT-Image-2.5 Flare 1,341
  4. 4 GPT-Image-2.6 Flare quality: auto, complexity: high 1,338
  5. 5 GPT-Image-2 1,315
  6. 6 Riverflow 2.5 Pro 1,286
  7. 7 Krea 3 ultra thinking 1,279
  8. 8 Reve 2.0 1,271
  9. 9 Reve 2.1 1,268
  10. 10 ChatGPT Image 1,268
  11. 11 Grok Imagine Image 2 Auto 1,254
  12. 12 Grok Imagine Image 2.0 medium 1,251
  13. 13 Muse Image 1,249
  14. 14 FLUX 3 Image Edit 1,247
  15. 15 Seedream 5.0 Pro Edit 1,239
  16. 16 MAI-Image-2.6 Flash 1,235
  17. 17 GPT-Image-1.5 1,228
  18. 18 Qwen-Image-3.0-Pro 1,228
  19. 19 MiniMax H3-Image 1,228
  20. 20 Gemini 3 Pro Image Preview 2K 1,227
  21. 21 MAI-Image-2.6 1,224
  22. 22 Gemini 3 Pro Image Preview 1,220
  23. 23 Ideogram 4.5 high 1,218
  24. 24 MAI-Image-2.5 1,202
  25. 25 Grok Imagine 1,177
  26. 26 Gemini 3.1 Flash-Lite Image (nano-banana-2-lite) 1,169
  27. 27 UNI-1.1 1,169
  28. 28 Seedream Lite 5.0 Edit 1,167
  29. 29 FLUX 3 Image Klein Edit 1,155
  30. 30 UNI-1.1 Max 1,146
  31. 31 Seedream 4.5 Edit 1,141
  32. 32 Wan 2.7 1,126
  33. 33 Wan 2.7 Pro 1,120
  34. 34 FLUX.2 Pro Edit 1,111
  35. 35 DreamImage 2.0 1,110
  36. 36 ImagineArt 2.0 Edit 1,102
  37. 37 Seedream 4.0 Edit 4K 1,097
  38. 38 Seedream 4.0 Edit 1,095
  39. 39 GPT-Image-1 1,081
  40. 40 FLUX.2 Klein 9B Edit 1,075
  41. 41 Reve Edit 1,068
  42. 42 Qwen-Image-Edit (2511) 1,064
  43. 43 Reve Edit Fast 1,045
  44. 44 P-Image-Edit 1,043
  45. 45 FLUX.2 Klein 4B Edit 1,032
  46. 46 FLUX.2 Klein 4B Distilled Edit 1,031
  47. 47 FLUX Kontext Max Edit 1,018
  48. 48 Boogu-Image-0.1 1,011
  49. 49 FLUX Kontext Pro Edit 1,002
  50. 50 Qwen-Image-Edit 1,001
  51. 51 HiDream O1 Image Edit 972
  52. 52 HiDream O1 Image Dev Edit 961
  53. 53 Gemini 2.0 Flash Edit 906
  54. 54 Recraft V3 Edit 787

Source: Official DesignArena Image Editing leaderboard

#EntryScoreSettingDateSource
1 GPT-Image-2.5 Sunburst 1,370 Official leaderboard
2 GPT-Image-2.6 Sunburst 1,349 quality: max, complexity: high Nov 2026 Official leaderboard
3 GPT-Image-2.5 Flare 1,341 Official leaderboard
4 GPT-Image-2.6 Flare 1,338 quality: auto, complexity: high Nov 2026 Official leaderboard
5 GPT-Image-2 1,315 Official leaderboard
6 Riverflow 2.5 Pro 1,286 Official leaderboard
7 Krea 3 1,279 ultra thinking Official leaderboard
8 Reve 2.0 1,271 Official leaderboard
9 Reve 2.1 1,268 Official leaderboard
10 ChatGPT Image 1,268 Official leaderboard
11 Grok Imagine Image 2 Auto 1,254 Official leaderboard
12 Grok Imagine Image 2.0 1,251 medium Official leaderboard
13 Muse Image 1,249 Official leaderboard
14 FLUX 3 Image Edit 1,247 Official leaderboard
15 Seedream 5.0 Pro Edit 1,239 Official leaderboard
16 MAI-Image-2.6 Flash 1,235 Official leaderboard
17 GPT-Image-1.5 1,228 Official leaderboard
18 Qwen-Image-3.0-Pro 1,228 Official leaderboard
19 MiniMax H3-Image 1,228 Official leaderboard
20 Gemini 3 Pro Image Preview 2K 1,227 Official leaderboard
21 MAI-Image-2.6 1,224 Official leaderboard
22 Gemini 3 Pro Image Preview 1,220 Official leaderboard
23 Ideogram 4.5 1,218 high Official leaderboard
24 MAI-Image-2.5 1,202 Official leaderboard
25 Grok Imagine 1,177 Official leaderboard
26 Gemini 3.1 Flash-Lite Image (nano-banana-2-lite) 1,169 Official leaderboard
27 UNI-1.1 1,169 Official leaderboard
28 Seedream Lite 5.0 Edit 1,167 Official leaderboard
29 FLUX 3 Image Klein Edit 1,155 Official leaderboard
30 UNI-1.1 Max 1,146 Official leaderboard
31 Seedream 4.5 Edit 1,141 Official leaderboard
32 Wan 2.7 1,126 Official leaderboard
33 Wan 2.7 Pro 1,120 Official leaderboard
34 FLUX.2 Pro Edit 1,111 Official leaderboard
35 DreamImage 2.0 1,110 Official leaderboard
36 ImagineArt 2.0 Edit 1,102 Official leaderboard
37 Seedream 4.0 Edit 4K 1,097 Official leaderboard
38 Seedream 4.0 Edit 1,095 Official leaderboard
39 GPT-Image-1 1,081 Official leaderboard
40 FLUX.2 Klein 9B Edit 1,075 Official leaderboard
41 Reve Edit 1,068 Official leaderboard
42 Qwen-Image-Edit (2511) 1,064 Official leaderboard
43 Reve Edit Fast 1,045 Official leaderboard
44 P-Image-Edit 1,043 Official leaderboard
45 FLUX.2 Klein 4B Edit 1,032 Official leaderboard
46 FLUX.2 Klein 4B Distilled Edit 1,031 Official leaderboard
47 FLUX Kontext Max Edit 1,018 Official leaderboard
48 Boogu-Image-0.1 1,011 Official leaderboard
49 FLUX Kontext Pro Edit 1,002 Official leaderboard
50 Qwen-Image-Edit 1,001 Official leaderboard
51 HiDream O1 Image Edit 972 Official leaderboard
52 HiDream O1 Image Dev Edit 961 Official leaderboard
53 Gemini 2.0 Flash Edit 906 Official leaderboard
54 Recraft V3 Edit 787 Official leaderboard

More effort adds a median 45 elo

Models the leaderboard ran at two or more effort levels. We don’t have enough cost data here to say what the extra effort costs.

Lowest effort to highest, light to dark
GPT-Image-2.6 Sunburst
quality: ma…
1,349
GPT-Image-2.6 Flare
quality: au… quality: au… quality: au…
1,338
Ideogram 4.5
high, very … high
1,218

Source: Official DesignArena Image Editing leaderboard

ModelEffortHarnessScoreCost per taskOutput tokens
GPT-Image-2.5 Sunburst 1,370
GPT-Image-2.6 Sunburst quality: max, complexity: high 1,349
GPT-Image-2.5 Flare 1,341
GPT-Image-2.6 Sunburst quality: max, complexity: high, quality: auto, complexity: high 1,340
GPT-Image-2.6 Flare quality: auto, complexity: high 1,338
GPT-Image-2 1,315
GPT-Image-2.6 Flare quality: auto, complexity: high, quality: low, complexity: medium 1,312
GPT-Image-2.6 Flare quality: auto, complexity: high, quality: low, complexity: low 1,293
Reve 2.0 1,271
Reve 2.1 1,268
Grok Imagine Image 2.0 medium 1,251
Muse Image 1,249
GPT-Image-1.5 1,228
Qwen-Image-3.0-Pro 1,228
MAI-Image-2.6 1,224
Ideogram 4.5 high 1,218
Qwen-Image-3.0-Pro 1,205
MAI-Image-2.5 1,202
Gemini 3.1 Flash-Lite Image (nano-banana-2-lite) 1,169
UNI-1.1 1,169
Ideogram 4.5 high, very low 1,168
UNI-1.1 Max 1,146
GPT-Image-1 1,081
Qwen-Image-Edit (2511) 1,064
Reve Edit Fast 1,045
Qwen-Image-Edit 1,001

Number one over time

Lab claims count from the day the card came out, independent runs from the model’s release.

OpenAI
Own leaderboard 1 leader
GPT-Image-2.6 Sunburst 1,349

Source: Official DesignArena Image Editing leaderboard

Scores over time

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

On the benchmark’s own leaderboard
1,280 1,300 1,320 1,340 1,360 2027 1,280 1,300 1,320 1,340 1,360 2027

Source: Official DesignArena Image Editing leaderboard

No lab model card reports it yet

We read the headline results of 522 launch posts from 52 labs, and none of them list this benchmark.

More in Vision

All 281 in Vision
  • MMMU In use 11.5K college-level multimodal questions across 30 subjects. Multiple-choice and open accuracy. GPT-5.1 85.4%
  • 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%
  • 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%
  • ChartQA In use Models answer human-written and generated questions about chart images that need visual and arithmetic reasoning. Scored by relaxed accuracy with a 5 percent numeric tolerance. Command A Vision 90.9%