Image model

GPT Image 2.5 Sunburst

OpenAI

Released: September 8, 2026
Data date: September 19, 2026

Access: API
Generator: gpt-image-2.5-sunburst
Benchmark configuration: Sunburst

No published category-wide ranking is available for this model.

Entries 1-10 of 10. Page 1 of 1.

Published benchmarks

Each chart compares models in the same documented test and highlights the current model.

Text to image

8 measurements

LMArena Text to Image: 3D modeling

Higher is better18 modelsGPT Image 2.5 Sunburst: Rank 1 of 18Data date: September 19, 2026

1.GPT Image 2.5 Sunburst1,404.71 points

2.GPT Image 21,361.09 points

3.Reve 2.11,275.82 points

4.Nano Banana 21,254.41 points

5.Seedream 5.0 Pro1,251.54 points

6.MAI-Image-2.51,248.55 points

7.Nano Banana Pro1,231.05 points

8.Nano Banana 2 Lite1,226.29 points

9.GPT Image 1.51,214.04 points

10.Qwen Image 2.0 Pro1,198.92 points

11.FLUX.2 Flex1,172.34 points

12.FLUX.2 Max1,161.4 points

13.FLUX.2 Pro1,147.77 points

14.Wan 2.6 Text to Image1,146.23 points

15.HiDream-O1-Image1,122.86 points

16.Krea 2 Large1,115.55 points

17.FLUX.2 Klein 9B1,078.7 points

18.Stable Diffusion 3.5 Large963.89 points

Scale: 0 points - 1,404.71 points

Source: LMArena

|

CC BY 4.0

gradually.ai

Methodology and context

Metric: arena-score
Scale: No upper bound is documented. The chart spans the recorded values and zero.
Participants: 18
Sample: 273

Source: LMArena

LMArena Text to Image: Art

Higher is better18 modelsGPT Image 2.5 Sunburst: Rank 1 of 18Data date: September 19, 2026

1.GPT Image 2.5 Sunburst1,390.12 points

2.GPT Image 21,365.89 points

3.Reve 2.11,287.42 points

4.MAI-Image-2.51,277.45 points

5.Seedream 5.0 Pro1,266.22 points

6.Nano Banana 2 Lite1,255.58 points

7.Nano Banana 21,241.43 points

8.GPT Image 1.51,224.79 points

9.Nano Banana Pro1,223.22 points

10.Qwen Image 2.0 Pro1,175.63 points

11.FLUX.2 Max1,169.99 points

12.FLUX.2 Pro1,162.17 points

13.Wan 2.6 Text to Image1,159.21 points

14.FLUX.2 Flex1,155.56 points

15.HiDream-O1-Image1,118.73 points

16.Krea 2 Large1,115.68 points

17.FLUX.2 Klein 9B1,078.94 points

18.Stable Diffusion 3.5 Large997.66 points

Scale: 0 points - 1,390.12 points

Source: LMArena

|

CC BY 4.0

gradually.ai

Methodology and context

Metric: arena-score
Scale: No upper bound is documented. The chart spans the recorded values and zero.
Participants: 18
Sample: 406

Source: LMArena

LMArena Text to Image: Cartoon

Higher is better18 modelsGPT Image 2.5 Sunburst: Rank 1 of 18Data date: September 19, 2026

1.GPT Image 2.5 Sunburst1,439.84 points

2.GPT Image 21,399.6 points

3.Reve 2.11,309.64 points

4.Seedream 5.0 Pro1,271.34 points

5.MAI-Image-2.51,269.58 points

6.Nano Banana 21,262.56 points

7.Nano Banana 2 Lite1,257.93 points

8.GPT Image 1.51,245.02 points

9.Nano Banana Pro1,242.36 points

10.Qwen Image 2.0 Pro1,191.07 points

11.FLUX.2 Max1,169.36 points

12.FLUX.2 Flex1,162.27 points

13.FLUX.2 Pro1,160.4 points

14.Wan 2.6 Text to Image1,146.21 points

15.HiDream-O1-Image1,120 points

16.Krea 2 Large1,112.78 points

17.FLUX.2 Klein 9B1,064.06 points

18.Stable Diffusion 3.5 Large945.68 points

Scale: 0 points - 1,439.84 points

Source: LMArena

|

CC BY 4.0

gradually.ai

Methodology and context

Metric: arena-score
Scale: No upper bound is documented. The chart spans the recorded values and zero.
Participants: 18
Sample: 1,198

Source: LMArena

LMArena Text to Image: Commercial design

Higher is better18 modelsGPT Image 2.5 Sunburst: Rank 1 of 18Data date: September 19, 2026

1.GPT Image 2.5 Sunburst1,418.82 points

2.GPT Image 21,389.09 points

3.Reve 2.11,314.48 points

4.Nano Banana 21,268.91 points

5.MAI-Image-2.51,260.92 points

6.Nano Banana 2 Lite1,250.51 points

7.Nano Banana Pro1,245.87 points

8.GPT Image 1.51,240.24 points

9.Seedream 5.0 Pro1,234.55 points

10.Qwen Image 2.0 Pro1,187.35 points

11.FLUX.2 Flex1,173.17 points

12.FLUX.2 Max1,164.74 points

13.FLUX.2 Pro1,155.35 points

14.Wan 2.6 Text to Image1,147.94 points

15.HiDream-O1-Image1,129.57 points

16.Krea 2 Large1,105.6 points

17.FLUX.2 Klein 9B1,069.62 points

18.Stable Diffusion 3.5 Large937.14 points

Scale: 0 points - 1,418.82 points

Source: LMArena

|

CC BY 4.0

gradually.ai

Methodology and context

Metric: arena-score
Scale: No upper bound is documented. The chart spans the recorded values and zero.
Participants: 18
Sample: 1,104

Source: LMArena

LMArena Text to Image: Overall

Higher is better18 modelsGPT Image 2.5 Sunburst: Rank 1 of 18Data date: September 19, 2026

1.GPT Image 2.5 Sunburst1,420.72 points

2.GPT Image 21,381.15 points

3.Reve 2.11,301.09 points

4.Nano Banana 21,260.68 points

5.Seedream 5.0 Pro1,257.1 points

6.MAI-Image-2.51,254.07 points

7.Nano Banana 2 Lite1,250.4 points

8.Nano Banana Pro1,246.24 points

9.GPT Image 1.51,239.16 points

10.Qwen Image 2.0 Pro1,191.16 points

11.FLUX.2 Max1,162.04 points

12.FLUX.2 Flex1,156.5 points

13.FLUX.2 Pro1,153.93 points

14.Wan 2.6 Text to Image1,136.59 points

15.HiDream-O1-Image1,116.96 points

16.Krea 2 Large1,108.26 points

17.FLUX.2 Klein 9B1,069.54 points

18.Stable Diffusion 3.5 Large938.34 points

Scale: 0 points - 1,420.72 points

Source: LMArena

|

CC BY 4.0

gradually.ai

Methodology and context

Metric: arena-score
Scale: No upper bound is documented. The chart spans the recorded values and zero.
Participants: 18
Sample: 3,149

Source: LMArena

LMArena Text to Image: Photorealism

Higher is better18 modelsGPT Image 2.5 Sunburst: Rank 1 of 18Data date: September 19, 2026

1.GPT Image 2.5 Sunburst1,423.22 points

2.GPT Image 21,380.61 points

3.Reve 2.11,302.39 points

4.Seedream 5.0 Pro1,280.52 points

5.Nano Banana 21,267.64 points

6.Nano Banana 2 Lite1,260.14 points

7.Nano Banana Pro1,259.84 points

8.MAI-Image-2.51,250.94 points

9.GPT Image 1.51,249.33 points

10.Qwen Image 2.0 Pro1,205.66 points

11.FLUX.2 Max1,161 points

12.FLUX.2 Pro1,152 points

13.FLUX.2 Flex1,146.87 points

14.Wan 2.6 Text to Image1,128.94 points

15.Krea 2 Large1,123.3 points

16.HiDream-O1-Image1,110.37 points

17.FLUX.2 Klein 9B1,090.2 points

18.Stable Diffusion 3.5 Large928.16 points

Scale: 0 points - 1,423.22 points

Source: LMArena

|

CC BY 4.0

gradually.ai

Methodology and context

Metric: arena-score
Scale: No upper bound is documented. The chart spans the recorded values and zero.
Participants: 18
Sample: 1,168

Source: LMArena

LMArena Text to Image: Portraits

Higher is better18 modelsGPT Image 2.5 Sunburst: Rank 1 of 18Data date: September 19, 2026

1.GPT Image 2.5 Sunburst1,450.12 points

2.GPT Image 21,427.41 points

3.Reve 2.11,316.2 points

4.Seedream 5.0 Pro1,294.77 points

5.Nano Banana 21,264.21 points

6.GPT Image 1.51,261.71 points

7.MAI-Image-2.51,257.76 points

8.Nano Banana Pro1,256.8 points

9.Nano Banana 2 Lite1,247.68 points

10.Qwen Image 2.0 Pro1,216 points

11.FLUX.2 Max1,157.11 points

12.FLUX.2 Pro1,144.64 points

13.FLUX.2 Flex1,144.63 points

14.Wan 2.6 Text to Image1,123.61 points

15.Krea 2 Large1,118.34 points

16.HiDream-O1-Image1,105.72 points

17.FLUX.2 Klein 9B1,091.96 points

18.Stable Diffusion 3.5 Large941.21 points

Scale: 0 points - 1,450.12 points

Source: LMArena

|

CC BY 4.0

gradually.ai

Methodology and context

Metric: arena-score
Scale: No upper bound is documented. The chart spans the recorded values and zero.
Participants: 18
Sample: 509

Source: LMArena

LMArena Text to Image: Text rendering

Higher is better18 modelsGPT Image 2.5 Sunburst: Rank 1 of 18Data date: September 19, 2026

1.GPT Image 2.5 Sunburst1,481.07 points

2.GPT Image 21,426.64 points

3.Reve 2.11,340.58 points

4.Nano Banana 21,291.89 points

5.MAI-Image-2.51,282.53 points

6.Nano Banana 2 Lite1,275.49 points

7.Nano Banana Pro1,269.94 points

8.Seedream 5.0 Pro1,254.64 points

9.GPT Image 1.51,253.44 points

10.Qwen Image 2.0 Pro1,201.08 points

11.FLUX.2 Flex1,171.99 points

12.FLUX.2 Max1,167.24 points

13.FLUX.2 Pro1,155.15 points

14.Wan 2.6 Text to Image1,148.6 points

15.HiDream-O1-Image1,134.27 points

16.Krea 2 Large1,111.57 points

17.FLUX.2 Klein 9B1,057.33 points

18.Stable Diffusion 3.5 Large921.53 points

Scale: 0 points - 1,481.07 points

Source: LMArena

|

CC BY 4.0

gradually.ai

Methodology and context

Metric: arena-score
Scale: No upper bound is documented. The chart spans the recorded values and zero.
Participants: 18
Sample: 1,125

Source: LMArena

Image editing

2 measurements

LMArena Image Editing: Multi-image editing

Higher is better11 modelsGPT Image 2.5 Sunburst: Rank 1 of 11Data date: September 19, 2026

1.GPT Image 2.5 Sunburst1,535.12 points

2.GPT Image 21,453.9 points

3.Seedream 5.0 Pro1,413.68 points

4.Nano Banana 21,367.77 points

5.Nano Banana Pro1,362.67 points

6.GPT Image 1.51,342.52 points

7.Nano Banana 2 Lite1,289.38 points

8.FLUX.2 Max1,250.29 points

9.FLUX.2 Pro1,240.56 points

10.FLUX.2 Flex1,234.15 points

11.FLUX.2 Klein 9B1,212.82 points

Scale: 0 points - 1,535.12 points

Source: LMArena

|

CC BY 4.0

gradually.ai

Methodology and context

Metric: arena-score
Scale: No upper bound is documented. The chart spans the recorded values and zero.
Participants: 11
Sample: 6,150

Source: LMArena

LMArena Image Editing: Overall

Higher is better15 modelsGPT Image 2.5 Sunburst: Rank 1 of 15Data date: September 19, 2026

1.GPT Image 2.5 Sunburst1,520.44 points

2.GPT Image 21,461.05 points

3.MAI-Image-2.51,400.21 points

4.Seedream 5.0 Pro1,394.48 points

5.Nano Banana Pro1,390.04 points

6.Nano Banana 21,387.49 points

7.Reve 2.11,375.31 points

8.GPT Image 1.51,370.19 points

9.Nano Banana 2 Lite1,314.3 points

10.Qwen Image 2.0 Pro1,303.65 points

11.HunyuanImage 3.0 Instruct1,302.41 points

12.FLUX.2 Max1,261.81 points

13.FLUX.2 Pro1,244.7 points

14.FLUX.2 Flex1,225.24 points

15.FLUX.2 Klein 9B1,224.42 points

Scale: 0 points - 1,520.44 points

Source: LMArena

|

CC BY 4.0

gradually.ai

Methodology and context

Metric: arena-score
Scale: No upper bound is documented. The chart spans the recorded values and zero.
Participants: 15
Sample: 6,704

Source: LMArena

Recorded measurement details

Every observation retains its source value and published test conditions.

LMArena Text to Image: 3D modeling (Show measurement, test conditions, and source)

Source value: 1404.713994277681
Score: 1404.713994277681
Metric: arena-score
Unit: score
Direction: Higher is better
Source rank: 1
Sample: 273
Confidence interval: 41.699436
Retrieved at: 2026-09-19

LMArena Text to Image: Art (Show measurement, test conditions, and source)

Source value: 1390.1199043561767
Score: 1390.1199043561767
Metric: arena-score
Unit: score
Direction: Higher is better
Source rank: 1
Sample: 406
Confidence interval: 31.10614
Retrieved at: 2026-09-19

LMArena Text to Image: Cartoon (Show measurement, test conditions, and source)

Source value: 1439.8408473789868
Score: 1439.8408473789868
Metric: arena-score
Unit: score
Direction: Higher is better
Source rank: 1
Sample: 1198
Confidence interval: 20.535447
Retrieved at: 2026-09-19

LMArena Text to Image: Commercial design (Show measurement, test conditions, and source)

Source value: 1418.8178164082756
Score: 1418.8178164082756
Metric: arena-score
Unit: score
Direction: Higher is better
Source rank: 1
Sample: 1104
Confidence interval: 20.951254
Retrieved at: 2026-09-19

LMArena Text to Image: Overall (Show measurement, test conditions, and source)

Source value: 1420.7153252330818
Score: 1420.7153252330818
Metric: arena-score
Unit: score
Direction: Higher is better
Source rank: 1
Sample: 3149
Confidence interval: 13.077556
Retrieved at: 2026-09-19

LMArena Text to Image: Photorealism (Show measurement, test conditions, and source)

Source value: 1423.2245542829473
Score: 1423.2245542829473
Metric: arena-score
Unit: score
Direction: Higher is better
Source rank: 1
Sample: 1168
Confidence interval: 20.322065
Retrieved at: 2026-09-19

LMArena Text to Image: Portraits (Show measurement, test conditions, and source)

Source value: 1450.124354091225
Score: 1450.124354091225
Metric: arena-score
Unit: score
Direction: Higher is better
Source rank: 2
Sample: 509
Confidence interval: 31.559462
Retrieved at: 2026-09-19

LMArena Text to Image: Text rendering (Show measurement, test conditions, and source)

Source value: 1481.0692970611149
Score: 1481.0692970611149
Metric: arena-score
Unit: score
Direction: Higher is better
Source rank: 1
Sample: 1125
Confidence interval: 22.656177
Retrieved at: 2026-09-19

LMArena Image Editing: Multi-image editing (Show measurement, test conditions, and source)

Source value: 1535.121048228931
Score: 1535.121048228931
Metric: arena-score
Unit: score
Direction: Higher is better
Source rank: 1
Sample: 6150
Confidence interval: 9.087024
Retrieved at: 2026-09-19

LMArena Image Editing: Overall (Show measurement, test conditions, and source)

Source value: 1520.443075793235
Score: 1520.443075793235
Metric: arena-score
Unit: score
Direction: Higher is better
Source rank: 1
Sample: 6704
Confidence interval: 9.205168
Retrieved at: 2026-09-19

Profile

Specifications and access

Published information about this model. Existing estimates are explicitly labeled.

Specification	Value and source
Weights	ProprietarySource
Access	APISource
Maximum output	3,840 × 2,160 pixelsSource
Benchmark configuration	SunburstSource
Generator	gpt-image-2.5-sunburstSource

Evidence

Sources and data date

Every statement links to its underlying documentation or leaderboard.

Type	Evidence and data date
Additional source	OpenAI
Additional source	LMArena (retrieved September 19, 2026)
Additional source	LMArena (retrieved September 19, 2026)