Benchmark Almanac

LMArena Image Edit Arena

GPT-Image-2.5 Sunburst 1,520 leads since November 2026
Not in model cards

Image editing models apply the same user edit request to the same image, anonymously and side by side. Community votes give Elo-style ratings.

Ratings are published in the image_edit config of the dataset.

Released
June 2025
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
59
Entries on its own leaderboard, 58 models we track

GPT-Image-2.5 Sunburst leads with 1,520

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. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders.

59 entries. Scroll sideways for the rest. Bars start at 800.

Bars start at 800.

  1. 1 GPT-Image-2.5 Sunburst 1,520
  2. 2 GPT-Image-2.5 Flare 1,484
  3. 3 GPT-Image-2.6 Sunburst medium 1,477
  4. 4 MAI-Image-2.7-Preview 1,468
  5. 5 GPT-Image-2 medium 1,461
  6. 6 GPT-Image-2.6 Flare medium 1,460
  7. 7 Grok Imagine Image 2.0 1,437
  8. 8 MAI-Image-2.6 1,428
  9. 9 Gemini Nano Banana 2.1 1,427
  10. 10 Krea 3 Ultra 1,411
  11. 11 MAI-Image-2.5 1,400
  12. 12 Muse Image 1,400
  13. 13 Seedream 5.0 Pro 1,391
  14. 14 Grok Imagine Image 1,391
  15. 15 Gemini 3 Pro Image (nano-banana-pro, 2K) 1,391
  16. 16 ChatGPT Image Latest high fidelity 1,387
  17. 17 Gemini 3.1 Flash Image (nano-banana-2, Web Search) 1,387
  18. 18 Gemini 3 Pro Image Preview (nano-banana-pro) 1,385
  19. 19 Reve 2.1 1,373
  20. 20 GPT-Image-1.5 high fidelity 1,367
  21. 21 Qwen-Image-2.1 1,363
  22. 22 Reve 2.0 1,359
  23. 23 Ideogram 4.5 1,346
  24. 24 UNI-1.1 Max 1,333
  25. 25 UNI-1.1 1,316
  26. 26 Gemini 3.1 Flash-Lite Image (nano-banana-2-lite) 1,313
  27. 27 Qwen-Image-2.0-Pro 1,303
  28. 28 Seedream 4.5 1,303
  29. 29 Wan 2.7 Image Pro 1,302
  30. 30 HunyuanImage 3.0 Instruct 1,301
  31. 31 Wan 2.7 Image 1,300
  32. 32 Seedream 5.0 Lite 1,294
  33. 33 Gemini 2.5 Flash Image Preview (nano-banana) 1,291
  34. 34 Seedream 4.0 1,267
  35. 35 FLUX.2 [max] 1,262
  36. 36 Reve v1.1 1,261
  37. 37 Kling Image O1 1,250
  38. 38 FLUX.2 [pro] 1,248
  39. 39 Qwen-Image-Edit 1,241
  40. 40 Reve v1 1,235
  41. 41 Qwen-Image-Edit (2511) 1,233
  42. 42 Wan 2.6 Image 1,231
  43. 43 FLUX.2 [klein] 9B 1,224
  44. 44 FLUX.2 [flex] 1,224
  45. 45 FLUX.2 [dev] 1,224
  46. 46 P-Image-Edit 1,210
  47. 47 Reve v1.1 Fast 1,207
  48. 48 Reve Edit Fast 1,198
  49. 49 FLUX.2 [klein] 4B 1,187
  50. 50 FLUX.1 Kontext [max] 1,182
  51. 51 Wan 2.5 Preview 1,180
  52. 52 FLUX.1 Kontext [pro] 1,177
  53. 53 FLUX.1 Kontext [dev] 1,150
  54. 54 SeedEdit 3.0 1,139
  55. 55 GPT-Image-1 1,139
  56. 56 GPT-Image-1 mini 1,124
  57. 57 Gemini 2.0 Flash Preview Image Generation 1,082
  58. 58 BAGEL 1,027
  59. 59 Step1X-Edit 996

Source: Official LMArena Image Edit Arena leaderboard

#EntryScoreSettingDateSource
1 GPT-Image-2.5 Sunburst 1,520 Nov 2026 Official leaderboard
2 GPT-Image-2.5 Flare 1,484 Nov 2026 Official leaderboard
3 GPT-Image-2.6 Sunburst 1,477 medium Nov 2026 Official leaderboard
4 MAI-Image-2.7-Preview 1,468 Nov 2026 Official leaderboard
5 GPT-Image-2 1,461 medium Nov 2026 Official leaderboard
6 GPT-Image-2.6 Flare 1,460 medium Nov 2026 Official leaderboard
7 Grok Imagine Image 2.0 1,437 Nov 2026 Official leaderboard
8 MAI-Image-2.6 1,428 Nov 2026 Official leaderboard
9 Gemini Nano Banana 2.1 1,427 Nov 2026 Official leaderboard
10 Krea 3 Ultra 1,411 Nov 2026 Official leaderboard
11 MAI-Image-2.5 1,400 Nov 2026 Official leaderboard
12 Muse Image 1,400 Nov 2026 Official leaderboard
13 Seedream 5.0 Pro 1,391 Nov 2026 Official leaderboard
14 Grok Imagine Image 1,391 May 2026 Official leaderboard
15 Gemini 3 Pro Image (nano-banana-pro, 2K) 1,391 Nov 2026 Official leaderboard
16 ChatGPT Image Latest 1,387 high fidelity Dec 2025 Official leaderboard
17 Gemini 3.1 Flash Image (nano-banana-2, Web Search) 1,387 Nov 2026 Official leaderboard
18 Gemini 3 Pro Image Preview (nano-banana-pro) 1,385 Nov 2026 Official leaderboard
19 Reve 2.1 1,373 Nov 2026 Official leaderboard
20 GPT-Image-1.5 1,367 high fidelity Nov 2026 Official leaderboard
21 Qwen-Image-2.1 1,363 Nov 2026 Official leaderboard
22 Reve 2.0 1,359 Nov 2026 Official leaderboard
23 Ideogram 4.5 1,346 Nov 2026 Official leaderboard
24 UNI-1.1 Max 1,333 Nov 2026 Official leaderboard
25 UNI-1.1 1,316 Nov 2026 Official leaderboard
26 Gemini 3.1 Flash-Lite Image (nano-banana-2-lite) 1,313 Nov 2026 Official leaderboard
27 Qwen-Image-2.0-Pro 1,303 Jun 2026 Official leaderboard
28 Seedream 4.5 1,303 Nov 2026 Official leaderboard
29 Wan 2.7 Image Pro 1,302 Nov 2026 Official leaderboard
30 HunyuanImage 3.0 Instruct 1,301 Nov 2026 Official leaderboard
31 Wan 2.7 Image 1,300 Nov 2026 Official leaderboard
32 Seedream 5.0 Lite 1,294 Nov 2026 Official leaderboard
33 Gemini 2.5 Flash Image Preview (nano-banana) 1,291 Nov 2026 Official leaderboard
34 Seedream 4.0 1,267 Nov 2026 Official leaderboard
35 FLUX.2 [max] 1,262 Nov 2026 Official leaderboard
36 Reve v1.1 1,261 Nov 2026 Official leaderboard
37 Kling Image O1 1,250 Nov 2026 Official leaderboard
38 FLUX.2 [pro] 1,248 Nov 2026 Official leaderboard
39 Qwen-Image-Edit 1,241 Nov 2026 Official leaderboard
40 Reve v1 1,235 Nov 2026 Official leaderboard
41 Qwen-Image-Edit (2511) 1,233 Nov 2026 Official leaderboard
42 Wan 2.6 Image 1,231 Nov 2026 Official leaderboard
43 FLUX.2 [klein] 9B 1,224 Nov 2026 Official leaderboard
44 FLUX.2 [flex] 1,224 Nov 2026 Official leaderboard
45 FLUX.2 [dev] 1,224 Nov 2026 Official leaderboard
46 P-Image-Edit 1,210 Nov 2026 Official leaderboard
47 Reve v1.1 Fast 1,207 Nov 2026 Official leaderboard
48 Reve Edit Fast 1,198 Nov 2026 Official leaderboard
49 FLUX.2 [klein] 4B 1,187 Nov 2026 Official leaderboard
50 FLUX.1 Kontext [max] 1,182 Nov 2026 Official leaderboard
51 Wan 2.5 Preview 1,180 Nov 2026 Official leaderboard
52 FLUX.1 Kontext [pro] 1,177 Nov 2026 Official leaderboard
53 FLUX.1 Kontext [dev] 1,150 Nov 2026 Official leaderboard
54 SeedEdit 3.0 1,139 Nov 2026 Official leaderboard
55 GPT-Image-1 1,139 Nov 2026 Official leaderboard
56 GPT-Image-1 mini 1,124 Nov 2026 Official leaderboard
57 Gemini 2.0 Flash Preview Image Generation 1,082 Nov 2026 Official leaderboard
58 BAGEL 1,027 Nov 2026 Official leaderboard
59 Step1X-Edit 996 Nov 2026 Official leaderboard

5 models have held number one on the official leaderboard since its release in June 2025

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

OpenAI xAI
Own leaderboard 5 leaders

Source: Official LMArena Image Edit Arena 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
800 1,000 1,200 1,400 1,600 2026 2027 800 1,000 1,200 1,400 1,600 2026 2027

Source: Official LMArena Image Edit Arena 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%