AI model directory
Explore AI Image Generation Models
Find an image model for the way you work. Explore text-to-image generation, reference-based creation, and editing in one directory.
Compare AI image models for text prompts, reference photos, and image editing. Review examples and choose a model for your next visual.
Browse AI Image Models
10 models
Google
OpenAI
ByteDance
Midjourney
Alibaba
Tongyi Lab
Xiaohongshu
xAI
Google
HiDream.aiCompare workflows
Compare Image Generation and Editing Workflows
Compare the workflow before choosing a model. Text prompts describe a new scene, while reference images provide a visual starting point for generation or editing.
ModelOutputText PromptReference InputsEditingOpen Model
See the Same Brief Across Image Models
Each task below keeps the input and full prompt fixed. Compare two recorded model runs, including the model version and output settings; the editing task also shows the shared reference image.
Product Composition
The same text input tests product placement, material detail, lighting, and the amount of usable campaign space.
Shared inputFull prompt
Full promptCreate a premium campaign still life featuring an unbranded translucent lavender perfume bottle and black cat-eye sunglasses on dark stone plinths. Add small purple flowers, a polished pale stone surface, soft blush-and-violet studio lighting, elegant shadows, and generous negative space. Wide landscape composition. No text, logos, or watermark.
GPT ImageModel output 1
- Version
- GPT Image 2
- Aspect ratio
- 3:1
- Output
- 1440 × 480
- Images
- 1
Nano BananaModel output 2
- Version
- Nano Banana 2
- Aspect ratio
- 3:1
- Output
- 2172 × 724
- Images
- 1
Poster Concept
The shared brief tests geometric layout, visual hierarchy, material rendering, and reserved copy space.
Shared inputFull prompt
Full promptCreate a premium photorealistic 3D poster mockup in a dark plum studio. Center one horizontal cream poster with bold plum and magenta geometric waves and a small chrome sphere. Surround it with restrained translucent acrylic and stone shapes under polished magenta studio light. Preserve generous blank cream space for future copy. Wide landscape composition. No readable text, logos, or watermark.
GPT ImageModel output 1
- Version
- GPT Image 2
- Aspect ratio
- 16:5
- Output
- 1440 × 450
- Images
- 1
Nano BananaModel output 2
- Version
- Nano Banana 2
- Aspect ratio
- 3:1
- Output
- 2172 × 724
- Images
- 1
Reference Image Edit
Both runs receive the same source image and edit instruction so you can compare subject preservation and background replacement.
Shared inputReference image
Full promptUse the uploaded image as the reference. Replace the entire environment with a bright warm-white daylight photography studio using a seamless plaster backdrop and soft shadows. Keep the cream boucle lounge chair unchanged as the sole hero product, preserving its position, scale, shape, upholstery texture, and camera angle. Remove all plants, archways, flowers, and colored garden light. Wide landscape composition. No text, logos, or watermark.
GPT ImageModel output 1
- Version
- GPT Image 2
- Input
- 1 reference image
- Aspect ratio
- 3:1
- Output
- 2172 × 724
Nano BananaModel output 2
- Version
- Nano Banana 2
- Input
- 1 reference image
- Aspect ratio
- 3:1
- Output
- 2172 × 724
Choose an Image Model for Your Task
Product Concepts
Compare backgrounds, lighting, and compositions for a product idea.
Product Concepts
Posters and Social Graphics
Review how each model handles layout, visual style, and the text in your brief.
Posters and Social Graphics
Editing a Reference Photo
Check how the model follows your requested change and which details it keeps from the input.
Editing a Reference PhotoFAQ
How do you choose the right AI model?
Direct answers about model inputs, workflows, settings, and availability.
What are AI image generation models?
AI image generation models create images from instructions such as a text prompt or reference photo. The models in this directory are available through WeShop AI workflows.
Which AI image model should I choose?
Choose based on your task: creating a new scene, exploring a layout, or editing an existing image. Compare examples with similar inputs and review the available controls.
Can I use a reference photo?
Reference-photo support is listed for each model. Open the model page to upload your image and review the available generation or editing settings.
Can I compare image models using the same brief?
Yes. Each of the three comparison tasks keeps one complete prompt fixed across both model outputs. The cards record the model version and settings, and the reference-edit task uses the same displayed source image for both runs.