Skip to content
Back to search

GLM-Image

GLM-Image: Auto-regressive for Dense-knowledge and High-fidelity Image Generation.

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked today
Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/zai-org--glm-image/badge.svg)](/listings/zai-org--glm-image)

HTML

<a href="/listings/zai-org--glm-image"><img src="/api/listings/zai-org--glm-image/badge.svg" alt="findsafeskills safety rating" /></a>

Install

i

Install instructions from the README

As written by the repo's author. findsafeskills has not verified these steps.

transformers + diffusers Pipeline

Install transformers and diffusers from source:

pip install git+https://github.com/huggingface/transformers.git
pip install git+https://github.com/huggingface/diffusers.git

+ Text to Image Generation

import torch
from diffusers.pipelines.glm_image import GlmImagePipeline

pipe = GlmImagePipeline.from_pretrained("zai-org/GLM-Image", torch_dtype=torch.bfloat16, device_map="cuda")
prompt = "A beautifully designed modern food magazine style dessert recipe illustration, themed around a raspberry mousse cake. The overall layout is clean and bright, divided into four main areas: the top left features a bold black title 'Raspberry Mousse Cake Recipe Guide', with a soft-lit close-up photo of the finished cake on the right, showcasing a light pink cake adorned with fresh raspberries and mint leaves; the bottom left contains an ingredient list section, titled 'Ingredients' in a simple font, listing 'Flour 150g', 'Eggs 3', 'Sugar 120g', 'Raspberry puree 200g', 'Gelatin sheets 10g', 'Whipping cream 300ml', and 'Fresh raspberries', each accompanied by minimalist line icons (like a flour bag, eggs, sugar jar, etc.); the bottom right displays four equally sized step boxes, each containing high-definition macro photos and corresponding instructions, arranged from top to bottom as follows: Step 1 shows a whisk whipping white foam (with the instruction 'Whip egg whites to stiff peaks'), Step 2 shows a red-and-white mixture being folded with a spatula (with the instruction 'Gently fold in the puree and batter'), Step 3 shows pink liquid being poured into a round mold (with the instruction 'Pour into mold and chill for 4 hours'), Step 4 shows the finished cake decorated with raspberries and mint leaves (with the instruction 'Decorate with raspberries and mint'); a light brown information bar runs along the bottom edge, with icons on the left representing 'Preparation time: 30 minutes', 'Cooking time: 20 minutes', and 'Servings: 8'. The overall color scheme is dominated by creamy white and light pink, with a subtle paper texture in the background, featuring compact and orderly text and image layout with clear information hierarchy."
image = pipe(
    prompt=prompt,
    height=32 * 32,
    width=36 * 32,
    num_inference_steps=50,
    guidance_scale=1.5,
    generator=torch.Generator(device="cuda").manual_seed(42),
).images[0]

image.save("output_t2i.png")

+ Image to Image Generation

import torch
from diffusers.pipelines.glm_image import GlmImagePipeline
from PIL import Image

pipe = GlmImagePipeline.from_pretrained("zai-org/GLM-Image", torch_dtype=torch.bfloat16, device_map="cuda")
image_path = "cond.jpg"
prompt = "Replace the background of the snow forest with an underground station featuring an automatic escalator."
image = Image.open(image_path).convert("RGB")
image = pipe(
    prompt=prompt,
    image=[image],  # can input multiple images for multi-image-to-image generation such as [image, image1]
    height=33 * 32, # Must set height even it is same as input image
    width=32 * 32, # Must set width even it is same as input image
    num_inference_steps=50,
    guidance_scale=1.5,
    generator=torch.Generator(device="cuda").manual_seed(42),
).images[0]

image.save("output_i2i.png")

SGLang Pipeline

Install transformers and diffusers from source:

pip install "sglang[diffusion] @ git+https://github.com/sgl-project/sglang.git#subdirectory=python"
pip install git+https://github.com/huggingface/transformers.git
pip install git+https://github.com/huggingface/diffusers.git

+ Text to Image Generation

sglang serve --model-path zai-org/GLM-Image

curl http://localhost:30000/v1/images/generations \
  -H "Content-Type: application/json" \
  -d '{
    "model": "zai-org/GLM-Image",
    "prompt": "a beautiful girl with glasses.",
    "n": 1,
    "response_format": "b64_json",
    "size": "1024x1024"

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

README.md · README · 465 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

  1. # GLM-Image
  2. <div align="center">
  3. <img src=resources/logo.svg width="40%"/>
  4. </div>
  5. <p align="center">
  6. 👋 Join our <a href="resources/WECHAT.md" target="_blank">WeChat</a> and <a href="https://discord.gg/8KFjEec7" target="_blank">Discord</a> community
  7. <br>
  8. 📖 Check out GLM-Image's <a href="https://z.ai/blog/glm-image" target="_blank">Technical Blog</a> and <a href="https://huggingface.co/zai-org/GLM-Image" target="_blank">🤗 Model Card</a>
  9. <br>
  10. 📍 Use GLM-Image's <a href="https://docs.z.ai/guides/image/glm-image" target="_blank">API</a>
  11. </p>
  12. <p align="center">
  13. <img src="resources/show_case.jpeg" alt="show_case" width="100%" />
  14. </p>
  15. ## Introduction
  16. GLM-Image is an image generation model adopts a hybrid autoregressive + diffusion decoder architecture. In general image generation quality, GLM‑Image aligns with mainstream latent diffusion approaches, but it shows significant advantages in text-rendering and knowledge‑intensive generation scenarios. It performs especially well in tasks requiring precise semantic understanding and complex information expression, while maintaining strong capabilities in high‑fidelity and fine‑grained detail generation. In addition to text‑to‑image generation, GLM‑Image also supports a rich set of image‑to‑image tasks including image editing, style transfer, identity‑preserving generation, and multi‑subject consistency.
  17. Model architecture: a hybrid autoregressive + diffusion decoder design.
  18. <p align="center">
  19. <img src="resources/architecture_1.jpeg" alt="architecture_1" width="100%" />
  20. </p>
  21. + Autoregressive generator: a 9B-parameter model initialized from [GLM-4-9B-0414](https://huggingface.co/zai-org/GLM-4-9B-0414), with an expanded vocabulary to incorporate visual tokens. The model first generates a compact encoding of approximately 256 tokens, then expands to 1K–4K tokens, corresponding to 1K–2K high-resolution image outputs.
  22. + Diffusion Decoder: a 7B-parameter decoder based on a single-stream DiT architecture for latent-space image decoding. It is equipped with a Glyph Encoder text module, significantly improving accurate text rendering within images.
  23. <p align="center">
  24. <img src="resources/architecture_2.jpeg" alt="architecture_2" width="70%" />
  25. </p>
  26. Post-training with decoupled reinforcement learning: the model introduces a fine-grained, modular feedback strategy using the GRPO algorithm, substantially enhancing both semantic understanding and visual detail quality.
  27. + Autoregressive module: provides low-frequency feedback signals focused on aesthetics and semantic alignment, improving instruction following and artistic expressiveness.
  28. + Decoder module: delivers high-frequency feedback targeting detail fidelity and text accuracy, resulting in highly realistic textures as well as more precise text rendering.
  29. GLM-Image supports both text-to-image and image-to-image generation within a single model.
  30. + Text-to-image: generates high-detail images from textual descriptions, with particularly strong performance in information-dense scenarios.
  31. + Image-to-image: supports a wide range of tasks, including image editing, style transfer, multi-subject consistency, and identity-preserving generation for people and objects.
  32. > You can find the full GLM-Image Model implementation in the [transformers](https://github.com/huggingface/transformers/tree/main/src/transformers/models/glm_image) and [diffusers](https://github.com/huggingface/diffusers/tree/main/src/diffusers/pipelines/glm_image) libraries here.
  33. ## Download
  34. | Model | Download Links |
  35. |-----------|------------------------------------------------------------------------------------------------------------------------------|
  36. | GLM-Image | [🤗 Hugging Face](https://huggingface.co/zai-org/GLM-Image)<br>[🤖 ModelScope](https://modelscope.cn/models/ZhipuAI/GLM-Image) |
  37. ## Showcase
  38. ### T2I with dense text and knowledge
  39. <p align="center">
  40. <img src="resources/show_case_t2i.jpeg" alt="show_case_t2i" width="100%" />
  41. </p>
  42. ### I2I
  43. <p align="center">
  44. <img src="resources/show_case_i2i.jpeg" alt="show_case_i2i" width="100%" />
  45. </p>
  46. ## Quick Start
  47. ### transformers + diffusers Pipeline
  48. Install transformers and diffusers from source:
  49. ```shell
  50. pip install git+https://github.com/huggingface/transformers.git
  51. pip install git+https://github.com/huggingface/diffusers.git
  52. ```
  53. + Text to Image Generation
  54. ```python
  55. import torch
  56. from diffusers.pipelines.glm_image import GlmImagePipeline
  57. pipe = GlmImagePipeline.from_pretrained("zai-org/GLM-Image", torch_dtype=torch.bfloat16, device_map="cuda")
  58. prompt = "A beautifully designed modern food magazine style dessert recipe illustration, themed around a raspberry mousse cake. The overall layout is clean and bright, divided into four main areas: the top left features a bold black title 'Raspberry Mousse Cake Recipe Guide', with a soft-lit close-up photo of the finished cake on the right, showcasing a light pink cake adorned with fresh raspberries and mint leaves; the bottom left contains an ingredient list section, titled 'Ingredients' in a simple font, listing 'Flour 150g', 'Eggs 3', 'Sugar 120g', 'Raspberry puree 200g', 'Gelatin sheets 10g', 'Whipping cream 300ml', and 'Fresh raspberries', each accompanied by minimalist line icons (like a flour bag, eggs, sugar jar, etc.); the bottom right displays four equally sized step boxes, each containing high-definition macro photos and corresponding instructions, arranged from top to bottom as follows: Step 1 shows a whisk whipping white foam (with the instruction 'Whip egg whites to stiff peaks'), Step 2 shows a red-and-white mixture being folded with a spatula (with the instruction 'Gently fold in the puree and batter'), Step 3 shows pink liquid being poured into a round mold (with the instruction 'Pour into mold and chill for 4 hours'), Step 4 shows the finished cake decorated with raspberries and mint leaves (with the instruction 'Decorate with raspberries and mint'); a light brown information bar runs along the bottom edge, with icons on the left representing 'Preparation time: 30 minutes', 'Cooking time: 20 minutes', and 'Servings: 8'. The overall color scheme is dominated by creamy white and light pink, with a subtle paper texture in the background, featuring compact and orderly text and image layout with clear information hierarchy."
  59. image = pipe(
  60. prompt=prompt,
  61. height=32 * 32,
  62. width=36 * 32,
  63. num_inference_steps=50,
  64. guidance_scale=1.5,
  65. generator=torch.Generator(device="cuda").manual_seed(42),
  66. ).images[0]
  67. image.save("output_t2i.png")
  68. ```
  69. + Image to Image Generation
  70. ```python
  71. import torch
  72. from diffusers.pipelines.glm_image import GlmImagePipeline
  73. from PIL import Image
  74. pipe = GlmImagePipeline.from_pretrained("zai-org/GLM-Image", torch_dtype=torch.bfloat16, device_map="cuda")
  75. image_path = "cond.jpg"
  76. prompt = "Replace the background of the snow forest with an underground station featuring an automatic escalator."
  77. image = Image.open(image_path).convert("RGB")
  78. image = pipe(
  79. prompt=prompt,
  80. image=[image], # can input multiple images for multi-image-to-image generation such as [image, image1]
  81. height=33 * 32, # Must set height even it is same as input image
  82. width=32 * 32, # Must set width even it is same as input image
  83. num_inference_steps=50,
  84. guidance_scale=1.5,
  85. generator=torch.Generator(device="cuda").manual_seed(42),
  86. ).images[0]
  87. image.save("output_i2i.png")
  88. ```
  89. ### SGLang Pipeline
  90. Install transformers and diffusers from source:
  91. ```
  92. pip install "sglang[diffusion] @ git+https://github.com/sgl-project/sglang.git#subdirectory=python"
  93. pip install git+https://github.com/huggingface/transformers.git
  94. pip install git+https://github.com/huggingface/diffusers.git
  95. ```
  96. + Text to Image Generation
  97. ```
  98. sglang serve --model-path zai-org/GLM-Image
  99. curl http://localhost:30000/v1/images/generations \
  100. -H "Content-Type: application/json" \
  101. -d '{
  102. "model": "zai-org/GLM-Image",
  103. "prompt": "a beautiful girl with glasses.",
  104. "n": 1,
  105. "response_format": "b64_json",
  106. "size": "1024x1024"
  107. }' | python3 -c "import sys, json, base64; open('output_t2i.png', 'wb').write(base64.b64decode(json.load(sys.stdin)['data'][0]['b64_json']))"
  108. ```
  109. + Image to Image Generation
  110. ```
  111. sglang serve --model-path zai-org/GLM-Image
  112. curl -s -X POST "http://localhost:30000/v1/images/edits" \
  113. -F "model=zai-org/GLM-Image" \
  114. -F "image=@cond.jpg" \
  115. -F "prompt=Replace the background of the snow forest with an underground station featuring an automatic escalator." \
  116. -F "response_format=b64_json" | python3 -c "import sys, json, base64; open('output_i2i.png', 'wb').write(base64.b64decode(json.load(sys.stdin)['data'][0]['b64_json']))"
  117. ```
  118. ### Note
  119. + Please ensure that all text intended to be rendered in the image is enclosed in quotation marks in the model input and We strongly recommend using GLM-4.7 to enhance prompts for higher image quality. Please check [script](examples/prompt_utils.py) for more details.
  120. + The AR model used in GLM‑Image is configured with `do_sample=True`, a temperature of `0.9`, and a topp of `0.75` by default. A higher temperature results in more diverse and rich outputs, but it can also lead to a certain decrease in output stability.
  121. + The target image resolution must be divisible by 32. Otherwise, it will throw an error.
  122. + Because the inference optimizations for this architecture are currently limited, the runtime cost is still relatively high. It requires either a single GPU with more than 80GB of memory, or a multi-GPU setup.
  123. + vLLM-Omni and SGLang (with AR speedup) support is currently being integrated — stay tuned.
  124. ### Inference Cost
  125. The following data was tested on a single `H100` with `diffusers` lib and can be used to estimate inference cost and time. This table may be updated after future optimizations.
  126. | Resolution | Batch Size | Type | End-to-End Time (s) | Peak VRAM (GB) | Note |
  127. |------------|------------|------|---------------------|----------------|--------------------------------------------------|
  128. | 2048×2048 | 1 | T2I | 252.59 | 45.08 | Fixed prompt, ~1000 tokens |
  129. | 1024×1024 | 1 | T2I | 64.34 | 37.82 | Fixed prompt, ~1000 tokens |
  130. | 1024×1024 | 4 | T2I | 108.22 | 52.09 | Fixed prompt, ~1000 tokens |
  131. | 512×512 | 1 | T2I | 27.30 | 34.26 | Fixed prompt, ~1000 tokens |
  132. | 512×512 | 4 | T2I | 39.34 | 37.83 | Fixed prompt, ~1000 tokens |
  133. | 2048×2048 | 1 | I2I | 274.51 | 45.08 | Reference image same size as target, ~100 tokens |
  134. | 1024×1024 | 1 | I2I | 58.64 | 37.83 | Reference image same size as target, ~100 tokens |
  135. ## Model Performance
  136. ### Text Rendering
  137. <div style="overflow-x: auto; margin-bottom: 16px;">
  138. <table style="border-collapse: collapse; width: 100%;">
  139. <thead>
  140. <tr>
  141. <th style="white-space: nowrap; padding: 8px; border: 1px solid #d0d7de; background-color: #f6f8fa;" rowspan="2">Model</th>
  142. <th style="white-space: nowrap; padding: 8px; border: 1px solid #d0d7de; background-color: #f6f8fa;" rowspan="2">Open Source</th>
  143. <th style="padding: 8px; border: 1px solid #d0d7de; background-color: #f6f8fa; text-align: center;" colspan="3">CVTG-2K</th>
  144. <th style="padding: 8px; border: 1px solid #d0d7de; background-color: #f6f8fa; text-align: center;" colspan="3">LongText-Bench</th>
  145. </tr>
  146. <tr>
  147. <th style="white-space: nowrap; padding: 8px; border: 1px solid #d0d7de; background-color: #f6f8fa; text-align: center;">Word Accuracy</th>
  148. <th style="padding: 8px; border: 1px solid #d0d7de; background-color: #f6f8fa; text-align: center;">NED</th>
  149. <th style="padding: 8px; border: 1px solid #d0d7de; background-color: #f6f8fa; text-align: center;">CLIPScore</th>
  150. <th style="padding: 8px; border: 1px solid #d0d7de; background-color: #f6f8fa; text-align: center;">AVG</th>
  151. <th style="padding: 8px; border: 1px solid #d0d7de; background-color: #f6f8fa; text-align: center;">EN</th>
  152. <th style="padding: 8px; border: 1px solid #d0d7de; background-color: #f6f8fa; text-align: center;">ZH</th>
  153. </tr>
  154. </thead>
  155. <tbody>
  156. <tr>
  157. <td style="padding: 8px; border: 1px solid #d0d7de;white-space:nowrap;">Seedream 4.5</td>
  158. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">✗</td>
  159. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.8990</td>
  160. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.9483</td>
  161. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;"><strong>0.8069</strong></td>
  162. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;"><strong>0.988</strong></td>
  163. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;"><strong>0.989</strong></td>
  164. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;"><strong>0.987</strong></td>
  165. </tr>
  166. <tr>
  167. <td style="padding: 8px; border: 1px solid #d0d7de;white-space:nowrap;">Seedream 4.0</td>
  168. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">✗</td>
  169. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.8451</td>
  170. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.9224</td>
  171. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.7975</td>
  172. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.924</td>
  173. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.921</td>
  174. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.926</td>
  175. </tr>
  176. <tr>
  177. <td style="padding: 8px; border: 1px solid #d0d7de;white-space:nowrap;">Nano Banana 2.0</td>
  178. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">✗</td>
  179. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.7788</td>
  180. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.8754</td>
  181. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.7372</td>
  182. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.965</td>
  183. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.981</td>
  184. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.949</td>
  185. </tr>
  186. <tr>
  187. <td style="padding: 8px; border: 1px solid #d0d7de;white-space:nowrap;">GPT Image 1 [High]</td>
  188. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">✗</td>
  189. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.8569</td>
  190. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.9478</td>
  191. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.7982</td>
  192. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.788</td>
  193. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.956</td>
  194. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.619</td>
  195. </tr>
  196. <tr>
  197. <td style="padding: 8px; border: 1px solid #d0d7de;white-space:nowrap;">Qwen-Image</td>
  198. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">✓</td>
  199. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.8288</td>
  200. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.9116</td>
  201. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.8017</td>
  202. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.945</td>
  203. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.943</td>
  204. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.946</td>
  205. </tr>
  206. <tr>
  207. <td style="padding: 8px; border: 1px solid #d0d7de;white-space:nowrap;">Qwen-Image-2512</td>
  208. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">✓</td>
  209. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.8604</td>
  210. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.9290</td>
  211. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.7819</td>
  212. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.961</td>
  213. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.956</td>
  214. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.965</td>
  215. </tr>
  216. <tr>
  217. <td style="padding: 8px; border: 1px solid #d0d7de;white-space:nowrap;">Z-Image</td>
  218. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">✓</td>
  219. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.8671</td>
  220. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.9367</td>
  221. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.7969</td>
  222. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.936</td>
  223. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.935</td>
  224. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.936</td>
  225. </tr>
  226. <tr>
  227. <td style="padding: 8px; border: 1px solid #d0d7de;white-space:nowrap;">Z-Image-Turbo</td>
  228. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">✓</td>
  229. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.8585</td>
  230. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.9281</td>
  231. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.8048</td>
  232. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.922</td>
  233. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.917</td>
  234. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.926</td>
  235. </tr>
  236. <tr>
  237. <td style="padding: 8px; border: 1px solid #d0d7de;white-space:nowrap;"><strong>GLM-Image</strong></td>
  238. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">✓</td>
  239. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;"><strong>0.9116</strong></td>
  240. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;"><strong>0.9557</strong></td>
  241. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.7877</td>
  242. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.966</td>
  243. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.952</td>
  244. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.979</td>
  245. </tr>
  246. </tbody>
  247. </table>
  248. </div>
  249. ### Text-to-Image
  250. <div style="overflow-x: auto; margin-bottom: 16px;">
  251. <table style="border-collapse: collapse; width: 100%;">
  252. <thead>
  253. <tr>
  254. <th style="white-space: nowrap; padding: 8px; border: 1px solid #d0d7de; background-color: #f6f8fa;" rowspan="2">Model</th>
  255. <th style="white-space: nowrap; padding: 8px; border: 1px solid #d0d7de; background-color: #f6f8fa;" rowspan="2">Open Source</th>
  256. <th style="padding: 8px; border: 1px solid #d0d7de; background-color: #f6f8fa; text-align: center;" colspan="2">OneIG-Bench</th>
  257. <th style="padding: 8px; border: 1px solid #d0d7de; background-color: #f6f8fa; text-align: center;" colspan="2">TIIF-Bench</th>
  258. <th style="white-space: nowrap; padding: 8px; border: 1px solid #d0d7de; background-color: #f6f8fa;" rowspan="2">DPG-Bench</th>
  259. </tr>
  260. <tr>
  261. <th style="white-space: nowrap; padding: 8px; border: 1px solid #d0d7de; background-color: #f6f8fa; text-align: center;">EN</th>
  262. <th style="white-space: nowrap; padding: 8px; border: 1px solid #d0d7de; background-color: #f6f8fa; text-align: center;">ZH</th>
  263. <th style="white-space: nowrap; padding: 8px; border: 1px solid #d0d7de; background-color: #f6f8fa; text-align: center;">short</th>
  264. <th style="white-space: nowrap; padding: 8px; border: 1px solid #d0d7de; background-color: #f6f8fa; text-align: center;">long</th>
  265. </tr>
  266. </thead>
  267. <tbody>
  268. <tr>
  269. <td style="padding: 8px; border: 1px solid #d0d7de; white-space:nowrap;">Seedream 4.5</td>
  270. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">✗</td>
  271. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.576</td>
  272. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.551</td>
  273. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">90.49</td>
  274. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;"><strong>88.52</strong></td>
  275. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;"><strong>88.63</strong></td>
  276. </tr>
  277. <tr>
  278. <td style="padding: 8px; border: 1px solid #d0d7de; white-space:nowrap;">Seedream 4.0</td>
  279. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">✗</td>
  280. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.576</td>
  281. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.553</td>
  282. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">90.45</td>
  283. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">88.08</td>
  284. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">88.54</td>
  285. </tr>
  286. <tr>
  287. <td style="padding: 8px; border: 1px solid #d0d7de; white-space:nowrap;">Nano Banana 2.0</td>
  288. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">✗</td>
  289. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;"><strong>0.578</strong></td>
  290. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;"><strong>0.567</strong></td>
  291. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;"><strong>91.00</strong></td>
  292. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">88.26</td>
  293. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">87.16</td>
  294. </tr>
  295. <tr>
  296. <td style="padding: 8px; border: 1px solid #d0d7de; white-space:nowrap;">GPT Image 1 [High]</td>
  297. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">✗</td>
  298. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.533</td>
  299. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.474</td>
  300. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">89.15</td>
  301. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">88.29</td>
  302. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">85.15</td>
  303. </tr>
  304. <tr>
  305. <td style="padding: 8px; border: 1px solid #d0d7de; white-space:nowrap;">DALL-E 3</td>
  306. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">✗</td>
  307. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">-</td>
  308. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">-</td>
  309. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">74.96</td>
  310. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">70.81</td>
  311. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">83.50</td>
  312. </tr>
  313. <tr>
  314. <td style="padding: 8px; border: 1px solid #d0d7de; white-space:nowrap;">Qwen-Image</td>
  315. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">✓</td>
  316. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.539</td>
  317. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.548</td>
  318. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">86.14</td>
  319. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">86.83</td>
  320. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">88.32</td>
  321. </tr>
  322. <tr>
  323. <td style="padding: 8px; border: 1px solid #d0d7de; white-space:nowrap;">Qwen-Image-2512</td>
  324. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">✓</td>
  325. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.530</td>
  326. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.515</td>
  327. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">83.24</td>
  328. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">84.93</td>
  329. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">87.20</td>
  330. </tr>
  331. <tr>
  332. <td style="padding: 8px; border: 1px solid #d0d7de; white-space:nowrap;">Z-Image</td>
  333. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">✓</td>
  334. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.546</td>
  335. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.535</td>
  336. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">80.20</td>
  337. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">83.01</td>
  338. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">88.14</td>
  339. </tr>
  340. <tr>
  341. <td style="padding: 8px; border: 1px solid #d0d7de; white-space:nowrap;">Z-Image-Turbo</td>
  342. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">✓</td>
  343. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.528</td>
  344. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.507</td>
  345. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">77.73</td>
  346. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">80.05</td>
  347. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">84.86</td>
  348. </tr>
  349. <tr>
  350. <td style="padding: 8px; border: 1px solid #d0d7de; white-space:nowrap;">FLUX.1 [Dev]</td>
  351. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">✓</td>
  352. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.434</td>
  353. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">-</td>
  354. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">71.09</td>
  355. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">71.78</td>
  356. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">83.52</td>
  357. </tr>
  358. <tr>
  359. <td style="padding: 8px; border: 1px solid #d0d7de; white-space:nowrap;">SD3 Medium</td>
  360. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">✓</td>
  361. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">-</td>
  362. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">-</td>
  363. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">67.46</td>
  364. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">66.09</td>
  365. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">84.08</td>
  366. </tr>
  367. <tr>
  368. <td style="padding: 8px; border: 1px solid #d0d7de; white-space:nowrap;">SD XL</td>
  369. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">✓</td>
  370. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.316</td>
  371. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">-</td>
  372. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">54.96</td>
  373. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">42.13</td>
  374. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">74.65</td>
  375. </tr>
  376. <tr>
  377. <td style="padding: 8px; border: 1px solid #d0d7de; white-space:nowrap;">BAGEL</td>
  378. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">✓</td>
  379. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.361</td>
  380. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.370</td>
  381. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">71.50</td>
  382. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">71.70</td>
  383. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">-</td>
  384. </tr>
  385. <tr>
  386. <td style="padding: 8px; border: 1px solid #d0d7de; white-space:nowrap;">Janus-Pro</td>
  387. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">✓</td>
  388. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.267</td>
  389. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.240</td>
  390. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">66.50</td>
  391. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">65.01</td>
  392. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">84.19</td>
  393. </tr>
  394. <tr>
  395. <td style="padding: 8px; border: 1px solid #d0d7de; white-space:nowrap;">Show-o2</td>
  396. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">✓</td>
  397. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.308</td>
  398. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">-</td>
  399. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">59.72</td>
  400. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">58.86</td>
  401. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">-</td>
  402. </tr>
  403. <tr>
  404. <td style="padding: 8px; border: 1px solid #d0d7de; white-space:nowrap;font-weight:bold;">GLM-Image</td>
  405. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">✓</td>
  406. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.528</td>
  407. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">0.511</td>
  408. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">81.01</td>
  409. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">81.02</td>
  410. <td style="padding: 8px; border: 1px solid #d0d7de; text-align: center;">84.78</td>
  411. </tr>
  412. </tbody>
  413. </table>
  414. </div>

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.