All General Coding Reasoning Multimodal Safety Specialized Agentic Language Long Context Speed Text-to-Image Image Editing Text-to-Video Image-to-Video
Image Editing Arena ELO (Artificial Analysis)69

Image Editing

Composite performance across 1 benchmarks · Normalized 0–100 · At least 1 scores required · Ranked by average performance · Top 50 of 69 eligible models

# Model Provider AA Img Edit Avg Coverage Price Value
1 GPT Image 2.5 Sunburst (max) OpenAI 1,199
100.0
1/1 $0.21/image 474.6
2 GPT Image 2.5 Flare (max) OpenAI 1,178
96.4
1/1 $0.21/image 457.5
3 MAI-Image-2.6 Microsoft 1,151
91.9
1/1 — —
4 Google: Nano Banana 2.1 Google 1,150
91.7
1/1 — —
5 MAI-Image-2.6-Flash Microsoft 1,142
90.4
1/1 — —
6 GPT Image 2 (high) OpenAI 1,133
88.8
1/1 $0.21/image 421.5
7 Muse Image Meta 1,126
87.6
1/1 $0.010/image 8760.0
8 Seedream 5.0 Pro ByteDance Seed 1,126
87.6
1/1 $0.045/image 1946.7
9 MAI-Image-2.5 Microsoft 1,125
87.5
1/1 — —
10 Google: Nano Banana 2 (Gemini 3.1 Flash Image) Google 1,121
86.8
1/1 $0.067/image 1291.7
11 MAI-Image-2.5-Pro Microsoft 1,117
86.1
1/1 — —
12 GPT Image 1.5 (high) OpenAI 1,116
86.0
1/1 $0.13/image 646.6
13 Grok Imagine Image 2.0 SpaceXAI 1,112
85.3
1/1 $0.040/image 2132.5
14 Google: Nano Banana Pro (Gemini 3 Pro Image) Google 1,109
84.8
1/1 $0.13/image 631.0
15 MAI-Image-2.5-Flash Microsoft 1,095
82.4
1/1 — —
16 Grok Imagine Image Quality SpaceXAI 1,091
81.7
1/1 $0.050/image 1634.0
17 HunyuanImage 3.5 (Preview) Tencent 1,088
81.2
1/1 — —
18 Qwen-Image-3.0-Pro Alibaba 1,087
81.0
1/1 $0.040/image 2025.0
19 Qwen-Image-2.1 Alibaba 1,080
79.9
1/1 — —
20 Luma UNI 1 Max Luma Labs 1,076
79.2
1/1 — —
21 Grok Imagine Image SpaceXAI 1,071
78.3
1/1 $0.020/image 3915.0
22 Ideogram 4.5 (High) Ideogram 1,065
77.3
1/1 — —
23 Qwen-Image-3.0 Alibaba 1,064
77.2
1/1 $0.030/image 2573.3
24 HunyuanImage 3.0 Instruct Tencent 1,063
77.0
1/1 — —
25 Seedream 5.0 Lite ByteDance Seed 1,061
76.6
1/1 $0.035/image 2188.6
26 Qwen Image 2.0 Pro (2026-04-22) Alibaba 1,057
76.0
1/1 $0.075/image 1013.3
27 Google: Nano Banana 2 Lite (Gemini 3.1 Flash Lite Image) Google 1,050
74.8
1/1 $0.034/image 2226.2
28 DreamImage 2.0 Newport AI 1,047
74.3
1/1 $0.045/image 1651.1
29 Wan 2.6 Image Alibaba 1,046
74.1
1/1 $0.030/image 2470.0
30 Wan 2.7 Pro Alibaba 1,045
73.9
1/1 $0.075/image 985.3
31 Luma UNI 1 Luma Labs 1,043
73.6
1/1 — —
32 Seedream 4.5 ByteDance Seed 1,041
73.3
1/1 $0.040/image 1832.5
33 Wan 2.7 Alibaba 1,038
72.8
1/1 $0.030/image 2426.7
34 HunyuanImage 3.0 (Tencent Cloud) Tencent 1,023
70.2
1/1 — —
35 FLUX.2 [dev] Flash Fal 1,021
69.9
1/1 — —
36 FLUX.2 [dev] Turbo Fal 1,012
68.4
1/1 — —
37 FLUX.2 [klein] 9B Black Forest Labs 1,011
68.2
1/1 $0.015/image 4546.7
38 FLUX.2 [pro] Black Forest Labs 1,009
67.9
1/1 $0.030/image 2263.3
39 FLUX.2 [flex] Black Forest Labs 1,009
67.9
1/1 $0.050/image 1358.0
40 Step Image Edit 2 Stepfun 1,005
67.2
1/1 $0.003/image 22400.0
41 FLUX.2 [dev] Black Forest Labs 1,000
66.3
1/1 — —
42 FLUX.2 [max] Black Forest Labs 995
65.5
1/1 $0.070/image 935.7
43 Qwen Image Edit Plus 2511 Alibaba 991
64.8
1/1 $0.030/image 2160.0
44 GPT Image 1 (high) OpenAI 991
64.8
1/1 $0.17/image 388.0
45 Qwen Image Edit Max 2601 Alibaba 985
63.8
1/1 $0.075/image 850.7
46 Google: Nano Banana (Gemini 2.5 Flash Image)Retiring Google 984
63.6
1/1 $0.039/image 1643.4
47 Reve V1 (December) Reve 982
63.3
1/1 — —
48 P-Image-Edit Pruna AI 981
63.1
1/1 $0.010/image 6310.0
49 Qwen Image 2.0 (2026-03-03) Alibaba 971
61.4
1/1 $0.035/image 1754.3
50 Qwen Image Edit Plus 2509 Alibaba 969
61.1
1/1 — —

GPT Image 2.5 Sunburst (max) leads on raw performance, but Step Image Edit 2 delivers 4620% more performance per dollar. That extra 32.8 points costs $0.006/image per point.

Access this data

JSON GET /api/v1 
[
  {
    "rank": 1,
    "model": "GPT Image 2.5 Sunburst (max)",
    "provider": "OpenAI",
    "avg_score": 100.0,
    "cost": 0.2107,
    "cost_unit": "image",
    "value": 474.6
  },
  {
    "rank": 2,
    "model": "GPT Image 2.5 Flare (max)",
    "provider": "OpenAI",
    "avg_score": 96.4,
    "cost": 0.2107,
    "cost_unit": "image",
    "value": 457.5
  },
  {
    "rank": 3,
    "model": "MAI-Image-2.6",
    "provider": "Microsoft",
    "avg_score": 91.9,
    "cost": null,
    "cost_unit": "image",
    "value": null
  }
]
Curl 
$ curl -s https://rightrank.ai/api/v1/rankings/image-editing.json | jq '.[:3]'
CLI brew install rightrank 
$ rightrank top image-editing --limit 5
$ rightrank compare claude-opus-4.6 gpt-5.4-pro --category image-editing
$ rightrank value image-editing  # best performance per dollar
Claude Code Skill Skill 
# In Claude Code, use the RightRank skill:
/rightrank best model for image-editing
/rightrank compare top 5 by value
/rightrank cheapest model scoring above 80 on image-editing