DreamWith is a one-stop AI workspace for image and video creation, officially positioned as completing AI image generation, AI video generation, and precise photo editing within one platform.
Purpose It solves the problem of creators needing to switch back and forth between multiple AI tools. Users can call more than 30 mainstream models, including GPT Image, Nano Banana, Flux Kontext, Veo, Kling, Seedance 2.0, and others, within the same workspace to complete image generation, video generation, and mask-based local retouching.
Main capabilities The platform provides three core functions:
- Image Generator: generates AI images based on prompts.
- Video Generator: generates AI videos.
- Mask Edit: upload an image, paint the area you want to edit, describe the modification, and then click edit to complete local replacement, repair, or redrawing.
How to use The official workflow for mask editing is as follows:
- Upload the source image: first prepare the original image that needs repair, replacement, or redrawing. PNG / JPG / WEBP formats are supported.
- Mask only the area to be changed: use the brush to paint closely along the target area so the model keeps the surrounding pixels stable.
- Describe the desired result: tell the model what should appear within the masked area, and make the description specific.
- View, download, or continue editing: you can use the first result as a base for further processing to gain more precise control.
The official guidance also provides some practical tips: for hair and transparent materials, paint a little further inside the edges; split large edits into two to three smaller processes; unless you want to redraw the whole image, avoid masking the entire image. In addition, the platform provides Prompt starters, so users can start with a template and then adjust it in the editor.
Use cases Suitable for creators who need to continuously complete text-to-image, text-to-video, and local photo retouching within the same interface, such as users who need to repeatedly fine-tune the same image or want to compare effects across different models.
Limitations The official materials do not specify details such as free quota, subscription pricing, number of generations, commercial licensing, or output resolution. This information is not yet provided.
