# Consistent Image Generation on GPT-4 generated character prompts

Source: [Consistent Image Generation on GPT-4 generated character prompts](<https://www.sota2.com/research/sota/consistent-image-generation-on-gpt-4-generated-character-prompts>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-02-26T06:06:22.415Z

Evaluation of image generation models focusing on maintaining character consistency across generated images based on prompts created by GPT-4, measured by Text-Image Similarity and Character Similarity metrics.

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

| Scope | Recorded value |
| --- | --- |
| Task | Consistent Image Generation |
| Dataset (including any stated split) | GPT-4 generated character prompts |
| Metric | Text-Image Similarity |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 3 of 3 |
| Coverage | All results currently returned for this benchmark |

Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol.

## Metrics

- [Text-Image Similarity](<https://www.sota2.com/research/sota/consistent-image-generation-on-gpt-4-generated-character-prompts.md?metric=Text-Image%20Similarity>): Higher is better; unit: PERCENT.

- [Character Similarity](<https://www.sota2.com/research/sota/consistent-image-generation-on-gpt-4-generated-character-prompts.md?metric=Character%20Similarity>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Text-Image Similarity | Result date | Report paper |
| --- | --- | --- | --- |
| StoryDiffusion | 65.86 | Not recorded | 68fa73161e209131cea5a71f |
| Photo Maker | 65.41 | Not recorded | 68fa73161e209131cea5a71f |
| IP-Adapter | 61.29 | Not recorded | 68fa73161e209131cea5a71f |

### StoryDiffusion

Text-Image Similarity: 65.86. Result ID: 6992beb93a6c3148491130f6.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73161e209131cea5a71f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73161e209131cea5a71f.md>)

[Implementation](<https://storydiffusion.github.io/>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### Photo Maker

Text-Image Similarity: 65.41. Result ID: 6992beb93a6c3148491130f5.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73161e209131cea5a71f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71651e209131cea512b1.md>)

[Implementation](<https://photo-maker.github.io/>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### IP-Adapter

Text-Image Similarity: 61.29. Result ID: 6992beb93a6c3148491130f4.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73161e209131cea5a71f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70591e209131cea4a798.md>)

[Implementation](<https://ip-adapter.github.io/>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.
