# Academic Illustration Generation on PaperBananaBench

Source: [Academic Illustration Generation on PaperBananaBench](<https://www.sota2.com/research/sota/academic-illustration-generation-on-paperbananabench>)

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

Data updated: 2026-06-01T07:55:00.323Z

Evaluation of models generating academic illustrations using the PaperBananaBench dataset, assessed using metrics such as Faithfulness, Conciseness, Readability, Aesthetic quality, and Overall score.

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 | Academic Illustration Generation |
| Dataset (including any stated split) | PaperBananaBench |
| Metric | Faithfulness |
| Unit | SCALAR |
| Direction | Higher is better |
| Results loaded | 17 of 17 |
| 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

- [Faithfulness](<https://www.sota2.com/research/sota/academic-illustration-generation-on-paperbananabench.md?metric=Faithfulness>): Higher is better; unit: SCALAR.

- [Conciseness](<https://www.sota2.com/research/sota/academic-illustration-generation-on-paperbananabench.md?metric=Conciseness>): Higher is better; unit: SCALAR.

- [Readability](<https://www.sota2.com/research/sota/academic-illustration-generation-on-paperbananabench.md?metric=Readability>): Higher is better; unit: SCALAR.

- [Aesthetic](<https://www.sota2.com/research/sota/academic-illustration-generation-on-paperbananabench.md?metric=Aesthetic>): Higher is better; unit: SCALAR.

- [Overall Score](<https://www.sota2.com/research/sota/academic-illustration-generation-on-paperbananabench.md?metric=Overall%20Score>): Higher is better; unit: SCALAR.

## Recorded results

| Method | Faithfulness | Result date | Report paper |
| --- | --- | --- | --- |
| Human | 50 | Not recorded | 69802fe1a8b01e76cc8e1887 |
| PaperBanana | 45.8 | Not recorded | 69802fe1a8b01e76cc8e1887 |
| Nano-Banana-Pro | 43 | Not recorded | 69802fe1a8b01e76cc8e1887 |
| Few-shot Nano-Banana-Pro | 41.6 | Not recorded | 69802fe1a8b01e76cc8e1887 |
| CRAFTER | 38.18 | Not recorded | 6a1d120cdbd7150506f83f85 |
| CRAFTER | 37.72 | Not recorded | 6a1d120cdbd7150506f83f85 |
| PaperBanana | 31.34 | Not recorded | 6a1d120cdbd7150506f83f85 |
| PaperBanana | 28.1 | Not recorded | 6a1d120cdbd7150506f83f85 |
| Nano Banana Pro | 25.43 | Not recorded | 6a1d120cdbd7150506f83f85 |
| PaperBanana | 16 | Not recorded | 69802fe1a8b01e76cc8e1887 |
| Nano Banana 2 | 15.07 | Not recorded | 6a1d120cdbd7150506f83f85 |
| AutoFigure | 12.33 | Not recorded | 6a1d120cdbd7150506f83f85 |
| GPT-Image-2\* | 8.42 | Not recorded | 6a1d120cdbd7150506f83f85 |
| Paper2Any | 6.5 | Not recorded | 69802fe1a8b01e76cc8e1887 |
| GPT-Image-1.5 | 4.5 | Not recorded | 69802fe1a8b01e76cc8e1887 |
| GLM-Image | 0 | Not recorded | 6a1d120cdbd7150506f83f85 |
| Qwen-Image | 0 | Not recorded | 6a1d120cdbd7150506f83f85 |

### Human

Faithfulness: 50. Result ID: 69965bc25698665ec4a15d98.

[Reporting paper](<https://www.sota2.com/research/paper/69802fe1a8b01e76cc8e1887.md>)

Recorded evaluation configuration: {"Framework Type":"Human Evaluation Reference"}. This record does not assert independent reproduction.

### PaperBanana

Faithfulness: 45.8. Result ID: 69965bc25698665ec4a15d97.

[Reporting paper](<https://www.sota2.com/research/paper/69802fe1a8b01e76cc8e1887.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69802fe1a8b01e76cc8e1887.md>)

[Implementation](<https://github.com/dwzhu-pku/PaperBanana>)

Recorded evaluation configuration: {"Framework Type":"Agentic Frameworks","Image Generation Model":"Nano-Banana-Pro"}. This record does not assert independent reproduction.

### Nano-Banana-Pro

Faithfulness: 43. Result ID: 69965bc25698665ec4a15d93.

[Reporting paper](<https://www.sota2.com/research/paper/69802fe1a8b01e76cc8e1887.md>)

Recorded evaluation configuration: {"Framework Type":"Vanilla Settings","Image Generation Model":"Nano-Banana-Pro"}. This record does not assert independent reproduction.

### Few-shot Nano-Banana-Pro

Faithfulness: 41.6. Result ID: 69965bc25698665ec4a15d94.

[Reporting paper](<https://www.sota2.com/research/paper/69802fe1a8b01e76cc8e1887.md>)

Recorded evaluation configuration: {"Framework Type":"Vanilla Settings","Image Generation Model":"Nano-Banana-Pro","Prompting Strategy":"Few-shot"}. This record does not assert independent reproduction.

### CRAFTER

Faithfulness: 38.18. Result ID: 6a1d3ac3fbc373592c162e1e.

[Reporting paper](<https://www.sota2.com/research/paper/6a1d120cdbd7150506f83f85.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a1d120cdbd7150506f83f85.md>)

[Implementation](<https://github.com/HaozheZhao/Crafter>)

Recorded evaluation configuration: {"Backbone":"Nano Banana 2","Type":"Agentic framework"}. This record does not assert independent reproduction.

### CRAFTER

Faithfulness: 37.72. Result ID: 6a1d3ac3fbc373592c162e1f.

[Reporting paper](<https://www.sota2.com/research/paper/6a1d120cdbd7150506f83f85.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a1d120cdbd7150506f83f85.md>)

[Implementation](<https://github.com/HaozheZhao/Crafter>)

Recorded evaluation configuration: {"Backbone":"Nano Banana Pro","Type":"Agentic framework"}. This record does not assert independent reproduction.

### PaperBanana

Faithfulness: 31.34. Result ID: 6a1d3ac3fbc373592c162e1d.

[Reporting paper](<https://www.sota2.com/research/paper/6a1d120cdbd7150506f83f85.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69802fe1a8b01e76cc8e1887.md>)

[Implementation](<https://github.com/dwzhu-pku/PaperBanana>)

Recorded evaluation configuration: {"Backbone":"Nano Banana Pro","Type":"Agentic framework"}. This record does not assert independent reproduction.

### PaperBanana

Faithfulness: 28.1. Result ID: 6a1d3ac3fbc373592c162e1c.

[Reporting paper](<https://www.sota2.com/research/paper/6a1d120cdbd7150506f83f85.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69802fe1a8b01e76cc8e1887.md>)

[Implementation](<https://github.com/dwzhu-pku/PaperBanana>)

Recorded evaluation configuration: {"Backbone":"Nano Banana 2","Type":"Agentic framework"}. This record does not assert independent reproduction.

### Nano Banana Pro

Faithfulness: 25.43. Result ID: 6a1d3ac3fbc373592c162e1a.

[Reporting paper](<https://www.sota2.com/research/paper/6a1d120cdbd7150506f83f85.md>)

Recorded evaluation configuration: {"Model Source":"Closed-source","Type":"Standalone generator"}. This record does not assert independent reproduction.

### PaperBanana

Faithfulness: 16. Result ID: 69965bc25698665ec4a15d96.

[Reporting paper](<https://www.sota2.com/research/paper/69802fe1a8b01e76cc8e1887.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69802fe1a8b01e76cc8e1887.md>)

[Implementation](<https://github.com/dwzhu-pku/PaperBanana>)

Recorded evaluation configuration: {"Framework Type":"Agentic Frameworks","Image Generation Model":"GPT-Image-1.5"}. This record does not assert independent reproduction.

### Nano Banana 2

Faithfulness: 15.07. Result ID: 6a1d3ac3fbc373592c162e19.

[Reporting paper](<https://www.sota2.com/research/paper/6a1d120cdbd7150506f83f85.md>)

Recorded evaluation configuration: {"Model Source":"Closed-source","Type":"Standalone generator"}. This record does not assert independent reproduction.

### AutoFigure

Faithfulness: 12.33. Result ID: 6a1d3ac3fbc373592c162e1b.

[Reporting paper](<https://www.sota2.com/research/paper/6a1d120cdbd7150506f83f85.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6982d2fea8b01e76cc8e9a5e.md>)

Recorded evaluation configuration: {"Backbone":"Nano Banana 2","Type":"Agentic framework"}. This record does not assert independent reproduction.

### GPT-Image-2\*

Faithfulness: 8.42. Result ID: 6a1d3ac3fbc373592c162e18.

[Reporting paper](<https://www.sota2.com/research/paper/6a1d120cdbd7150506f83f85.md>)

Recorded evaluation configuration: {"Model Source":"Closed-source","Type":"Standalone generator"}. This record does not assert independent reproduction.

### Paper2Any

Faithfulness: 6.5. Result ID: 69965bc25698665ec4a15d95.

[Reporting paper](<https://www.sota2.com/research/paper/69802fe1a8b01e76cc8e1887.md>)

Recorded evaluation configuration: {"Framework Type":"Agentic Frameworks","Image Generation Model":"Nano-Banana-Pro"}. This record does not assert independent reproduction.

### GPT-Image-1.5

Faithfulness: 4.5. Result ID: 69965bc25698665ec4a15d92.

[Reporting paper](<https://www.sota2.com/research/paper/69802fe1a8b01e76cc8e1887.md>)

Recorded evaluation configuration: {"Framework Type":"Vanilla Settings","Image Generation Model":"GPT-Image-1.5"}. This record does not assert independent reproduction.

### GLM-Image

Faithfulness: 0. Result ID: 6a1d3ac3fbc373592c162e16.

[Reporting paper](<https://www.sota2.com/research/paper/6a1d120cdbd7150506f83f85.md>)

Recorded evaluation configuration: {"Model Source":"Open-source","Type":"Standalone generator"}. This record does not assert independent reproduction.

### Qwen-Image

Faithfulness: 0. Result ID: 6a1d3ac3fbc373592c162e17.

[Reporting paper](<https://www.sota2.com/research/paper/6a1d120cdbd7150506f83f85.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a6a1e209131cea83137.md>)

Recorded evaluation configuration: {"Model Source":"Open-source","Type":"Standalone generator"}. This record does not assert independent reproduction.

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