ChatWise Docs

English

中文

English

中文

Appearance

Image Generation ​

ChatWise integrates powerful text-to-image generation capabilities, allowing you to create and edit images directly within your chats.

Getting Started ​

Enable image generation:

enable

After enabling image generation, click the arrow button to switch model:

model

Make sure you have set an API key for the provider of choice.

Now you can ask the model to generate an image:

prompt
generate a realistic magazine cover image for "CUTE ANIMALS", featuring three ridiculously cute pets decked out in stylish accessories, complete with convincing, eye-catching headlines and teaser text.

Editing Image ​

Currently only gpt-image-1 can edit images and keep images consistent.

Custom Providers ​

You can also use models from custom providers (OpenAI compatible), for example you can use OpenAI images models from AiHubMix:

custom providers

You also need to set the model type to Image:

Then the image model will be available:

Supported Image Models ​

Currently only support:

  • OpenAI: dall-e-3, gpt-image-1
  • xAI: grok-2-image
  • Images models from custom providers

Notes ​

  • For Gemini 2.0 Flash (Image Gen) googleai-gemini-2.0-flash-exp you don't need to enable this feature, since this model natively supports image generation.
  • If you can't access gpt-image-1 from OpenAI, try alternative third-party provider like AiHubMix.