Acropolis

LMArena Image Edit

Arena human-vote leaderboard: rating with confidence interval, vote count and rank. One row per model and category.

Image generationVision
Independent. Run by an operator independent of the model vendors.LMArena by ArenaVersion 1Data as of 17 Nov 2026

No filters. Showing 48 of 48 rows.

  1. 1
    OpenAIDossier

    Arena rating

    GPT Image 2.5 Sunburst, Arena rating: 1,519 (scale 950.0 to 1518.6), 95% confidence interval 1514.0 to 1523.21,519(1514.0–1523.2)
  2. 2
    OpenAIDossier

    Arena rating

    GPT Image 2.5 Flare, Arena rating: 1,484 (scale 950.0 to 1518.6), 95% confidence interval 1479.7 to 1488.41,484(1479.7–1488.4)
  3. 3
    OpenAIDossier

    medium

    Arena rating

    GPT Image 2.6 Sunburst, Arena rating: 1,476 (scale 950.0 to 1518.6), 95% confidence interval 1465.9 to 1487.01,476(1465.9–1487.0)
  4. 4
    OpenAIDossier

    medium

    Arena rating

    GPT Image 2, Arena rating: 1,462 (scale 950.0 to 1518.6), 95% confidence interval 1458.6 to 1464.61,462(1458.6–1464.6)
  5. 5
    OpenAIDossier

    medium

    Arena rating

    GPT Image 2.6 Flare, Arena rating: 1,460 (scale 950.0 to 1518.6), 95% confidence interval 1449.8 to 1470.61,460(1449.8–1470.6)
  6. 6

    Arena rating

    Grok Imagine Image 2.0, Arena rating: 1,437 (scale 950.0 to 1518.6), 95% confidence interval 1431.4 to 1443.01,437(1431.4–1443.0)
  7. 7
    MicrosoftDossier

    Arena rating

    MAI-Image-2.6, Arena rating: 1,428 (scale 950.0 to 1518.6), 95% confidence interval 1422.9 to 1432.21,428(1422.9–1432.2)
  8. 8
    GoogleDossier

    Arena rating

    Nano Banana 2.1, Arena rating: 1,427 (scale 950.0 to 1518.6), 95% confidence interval 1421.3 to 1433.21,427(1421.3–1433.2)
  9. 9

    Arena rating

    Grok Imagine Image 2.0, Arena rating: 1,418 (scale 950.0 to 1518.6), 95% confidence interval 1413.7 to 1422.11,418(1413.7–1422.1)
  10. 10

    Arena rating

    Muse Image 1.0, Arena rating: 1,401 (scale 950.0 to 1518.6), 95% confidence interval 1397.0 to 1404.51,401(1397.0–1404.5)
  11. 11
    MicrosoftDossier

    Arena rating

    MAI-Image-2.5, Arena rating: 1,400 (scale 950.0 to 1518.6), 95% confidence interval 1396.6 to 1404.11,400(1396.6–1404.1)
  12. 12
    ByteDanceDossier

    Arena rating

    Seedream 5.0 Pro, Arena rating: 1,392 (scale 950.0 to 1518.6), 95% confidence interval 1388.6 to 1395.01,392(1388.6–1395.0)
  13. 13

    quality

    Arena rating

    Grok Imagine Image, Arena rating: 1,391 (scale 950.0 to 1518.6), 95% confidence interval 1385.6 to 1396.91,391(1385.6–1396.9)
  14. 14
    GoogleDossier

    nano-banana-pro

    Arena rating

    Nano Banana Pro (Gemini 3 Pro Image), Arena rating: 1,391 (scale 950.0 to 1518.6), 95% confidence interval 1388.2 to 1393.31,391(1388.2–1393.3)
  15. 15
    GoogleDossier

    web-search

    Arena rating

    Nano Banana 2 (Gemini 3.1 Flash Image), Arena rating: 1,387 (scale 950.0 to 1518.6), 95% confidence interval 1383.8 to 1390.01,387(1383.8–1390.0)
  16. 16
    GoogleDossier

    nano-banana-pro

    Arena rating

    Nano Banana Pro (Gemini 3 Pro Image), Arena rating: 1,385 (scale 950.0 to 1518.6), 95% confidence interval 1381.9 to 1387.71,385(1381.9–1387.7)
  17. 17

    Arena rating

    Reve 2.1, Arena rating: 1,373 (scale 950.0 to 1518.6), 95% confidence interval 1366.6 to 1378.91,373(1366.6–1378.9)
  18. 18
    OpenAIDossier

    high-fidelity

    Arena rating

    GPT Image 1.5, Arena rating: 1,367 (scale 950.0 to 1518.6), 95% confidence interval 1364.6 to 1369.81,367(1364.6–1369.8)
  19. 19
    AlibabaDossier

    Arena rating

    Qwen Image 2.1, Arena rating: 1,363 (scale 950.0 to 1518.6), 95% confidence interval 1357.3 to 1369.61,363(1357.3–1369.6)
  20. 20

    Arena rating

    Reve 2.0, Arena rating: 1,359 (scale 950.0 to 1518.6), 95% confidence interval 1352.7 to 1365.71,359(1352.7–1365.7)
  21. 21
    MiniMaxDossier

    Arena rating

    H3-Image, Arena rating: 1,359 (scale 950.0 to 1518.6), 95% confidence interval 1349.9 to 1367.91,359(1349.9–1367.9)
  22. 22
    Luma LabsDossier

    Arena rating

    Luma Uni-1.1 Max, Arena rating: 1,333 (scale 950.0 to 1518.6), 95% confidence interval 1328.3 to 1338.21,333(1328.3–1338.2)
  23. 23

    Arena rating

    Grok Imagine Image, Arena rating: 1,328 (scale 950.0 to 1518.6), 95% confidence interval 1326.0 to 1330.81,328(1326.0–1330.8)
  24. 24
    Luma LabsDossier

    Arena rating

    Luma Uni-1.1, Arena rating: 1,316 (scale 950.0 to 1518.6), 95% confidence interval 1312.5 to 1318.91,316(1312.5–1318.9)
  25. 25
    GoogleDossier

    Arena rating

    Nano Banana 2 Lite (Gemini 3.1 Flash Lite Image), Arena rating: 1,313 (scale 950.0 to 1518.6), 95% confidence interval 1308.0 to 1317.61,313(1308.0–1317.6)
  26. 26
    AlibabaDossier

    Arena rating

    Qwen Image 2.0 Pro, Arena rating: 1,303 (scale 950.0 to 1518.6), 95% confidence interval 1298.0 to 1307.41,303(1298.0–1307.4)
  27. 27
    ByteDanceDossier

    Arena rating

    Seedream 4.5, Arena rating: 1,303 (scale 950.0 to 1518.6), 95% confidence interval 1300.3 to 1304.71,303(1300.3–1304.7)
  28. 28
    AlibabaDossier

    Arena rating

    Wan 2.7 Pro, Arena rating: 1,302 (scale 950.0 to 1518.6), 95% confidence interval 1298.0 to 1306.31,302(1298.0–1306.3)
  29. 29
    AlibabaDossier

    Arena rating

    Wan 2.7, Arena rating: 1,301 (scale 950.0 to 1518.6), 95% confidence interval 1296.9 to 1304.71,301(1296.9–1304.7)
  30. 30
    ByteDanceDossier

    Arena rating

    Seedream 5.0 Lite, Arena rating: 1,293 (scale 950.0 to 1518.6), 95% confidence interval 1290.8 to 1295.91,293(1290.8–1295.9)
  31. 31
    GoogleDossier

    nano-banana

    Arena rating

    Gemini 2.5 Flash Image Preview, Arena rating: 1,292 (scale 950.0 to 1518.6), 95% confidence interval 1289.5 to 1293.61,292(1289.5–1293.6)
  32. 32
    ByteDanceDossier

    Arena rating

    Seedream 4.0, Arena rating: 1,267 (scale 950.0 to 1518.6), 95% confidence interval 1260.1 to 1274.01,267(1260.1–1274.0)
  33. 33
    Black Forest LabsDossier

    Arena rating

    FLUX.2 [max], Arena rating: 1,262 (scale 950.0 to 1518.6), 95% confidence interval 1258.5 to 1264.91,262(1258.5–1264.9)
  34. 34
    Black Forest LabsDossier

    Arena rating

    FLUX.2 [pro], Arena rating: 1,247 (scale 950.0 to 1518.6), 95% confidence interval 1244.9 to 1250.01,247(1244.9–1250.0)
  35. 35
    Black Forest LabsDossier

    Arena rating

    FLUX.2 [klein] 9B, Arena rating: 1,224 (scale 950.0 to 1518.6), 95% confidence interval 1221.5 to 1227.31,224(1221.5–1227.3)
  36. 36
    Black Forest LabsDossier

    Arena rating

    FLUX.2 [flex], Arena rating: 1,224 (scale 950.0 to 1518.6), 95% confidence interval 1221.3 to 1227.61,224(1221.3–1227.6)
  37. 37
    Black Forest LabsDossier

    Arena rating

    FLUX.2 [dev], Arena rating: 1,224 (scale 950.0 to 1518.6), 95% confidence interval 1219.9 to 1228.01,224(1219.9–1228.0)
  38. 38
    ByteDanceDossier

    high-res · fal

    Arena rating

    Seedream 4.0, Arena rating: 1,216 (scale 950.0 to 1518.6), 95% confidence interval 1212.9 to 1218.51,216(1212.9–1218.5)
  39. 39
    ByteDanceDossier

    fal

    Arena rating

    Seedream 4.0, Arena rating: 1,209 (scale 950.0 to 1518.6), 95% confidence interval 1203.2 to 1215.51,209(1203.2–1215.5)
  40. 40
    Black Forest LabsDossier

    Arena rating

    FLUX.2 [klein] 4B, Arena rating: 1,187 (scale 950.0 to 1518.6), 95% confidence interval 1184.3 to 1190.11,187(1184.3–1190.1)
  41. 41
    Black Forest LabsDossier

    Arena rating

    FLUX.1 Kontext [max], Arena rating: 1,182 (scale 950.0 to 1518.6), 95% confidence interval 1179.3 to 1184.91,182(1179.3–1184.9)
  42. 42
    AlibabaDossier

    Arena rating

    Wan 2.5 Preview, Arena rating: 1,180 (scale 950.0 to 1518.6), 95% confidence interval 1177.1 to 1182.51,180(1177.1–1182.5)
  43. 43
    Black Forest LabsDossier

    Arena rating

    FLUX.1 Kontext [pro], Arena rating: 1,177 (scale 950.0 to 1518.6), 95% confidence interval 1174.6 to 1180.01,177(1174.6–1180.0)
  44. 44
    Black Forest LabsDossier

    Arena rating

    FLUX.1 Kontext [dev], Arena rating: 1,150 (scale 950.0 to 1518.6), 95% confidence interval 1146.7 to 1153.71,150(1146.7–1153.7)
  45. 45
    OpenAIDossier

    Arena rating

    GPT Image 1, Arena rating: 1,139 (scale 950.0 to 1518.6), 95% confidence interval 1136.0 to 1141.51,139(1136.0–1141.5)
  46. 46
    OpenAIDossier

    Arena rating

    GPT Image 1 Mini, Arena rating: 1,124 (scale 950.0 to 1518.6), 95% confidence interval 1121.0 to 1127.01,124(1121.0–1127.0)
  47. 47
    GoogleDossier

    Arena rating

    Gemini 2.0 Flash Preview Image Generation, Arena rating: 1,082 (scale 950.0 to 1518.6), 95% confidence interval 1079.4 to 1084.31,082(1079.4–1084.3)
  48. 48
    ByteDanceDossier

    Arena rating

    BAGEL, Arena rating: 1,027 (scale 950.0 to 1518.6), 95% confidence interval 1020.8 to 1032.31,027(1020.8–1032.3)
Unmatched upstream rows excluded from this board (16)

These upstream rows could not be mapped cleanly to the canonical model registry or reasoning effort enum.

Raw nameParsed modelReasonDetail
MAI-Image-2.7-PreviewMAI-Image-2.7-Previewno-aliasrating
Krea 3 UltraKrea 3 Ultrano-aliasrating
ChatGPT Image Latest (High Fidelity, 20251216)ChatGPT Image Latestno-aliasrating
Ideogram 4.5Ideogram 4.5no-aliasrating
HunyuanImage 3.0 InstructHunyuanImage 3.0 Instructno-aliasrating
Reve v1.1Reve v1.1no-aliasrating
Kling Image O1Kling Image O1no-aliasrating
Qwen-Image-EditQwen-Image-Editno-aliasrating
Reve v1Reve v1no-aliasrating
Qwen-Image-Edit (2511)Qwen-Image-Edit (2511)no-aliasrating
Wan 2.6 ImageWan 2.6 Imageno-aliasrating
P-Image-EditP-Image-Editno-aliasrating
Reve v1.1 FastReve v1.1 Fastno-aliasrating
Reve Edit FastReve Edit Fastno-aliasrating
SeedEdit 3.0SeedEdit 3.0no-aliasrating
Step1X-EditStep1X-Editno-aliasrating