CGTrader AI

CGDream's image tools move into CGTrader AI

CGTrader AI connects image generation with 3D asset creation and refinement. The former CGDream address now leads to this service, and the official page explains that CGDream is retiring as a standalone product. This listing covers the current AI tool family, rather than the separate marketplace for buying models.

The workflow can suit a concept artist developing a visual direction, a game creator preparing a prototype prop or a designer exploring how a form looks from different viewpoints. It brings generated reference images and 3D operations into the same product context, while the creator remains responsible for deciding whether the result fits a production pipeline.

Image generation, editing and 3D controls

The image tools accept text, image or 3D inputs. A 3D viewer provides a way to control the viewpoint of a visual, while styles and filters change its treatment. Inpainting targets a selected part of an image, and an upscaler helps prepare a larger version. Those operations have different purposes: changing a background is not the same as changing an object's geometry.

The connected 3D generator provides text-to-3D and image-to-3D paths. Its enhancement tools address topology, textures, segmentation and animation. The official page also describes editing your own models and eligible marketplace models. Eligibility matters: purchasing or downloading an asset does not automatically establish permission for every modification or redistribution.

Example: develop a stylized game prop

Suppose you need a small lantern for a game prototype. Begin with a short description of its silhouette, materials and intended style. Generate several reference directions and select one whose handle, base and light housing are clear. If a local image change is needed, use inpainting to adjust that area while comparing the rest of the composition with the selected reference.

Use the reference in the image-to-3D workflow, then inspect the result from several angles. The front view may look persuasive while the back, underside or openings need correction. Review the mesh and texture separately: a painted dark patch can suggest a hole without creating usable geometry.

Try the relevant enhancement controls, then check the asset in the application where it will be used. Confirm scale, orientation, shading and the detail budget for the scene. Keep the reference, revised model and notes together so another collaborator can understand which parts are provisional. This is an exploration workflow; a generated prop should not be treated as a dimensionally accurate engineering design.

Migration and production limitations

The migration page explicitly says some former CGDream options, including Flux Dev, Juggernaut and LoRA support, are not available in the current service. A workflow that depended on one of those options needs reassessment rather than an assumption that a saved recipe transfers unchanged.

Generated surfaces, textures and topology still need inspection against the destination's requirements. A concept image, a printable object and an animated game asset impose different constraints. Verify the required export format and editing controls in the account before committing a project to the service; the public overview does not establish every downstream application's compatibility.

Free generations and subscription access

CGTrader's official pages offer a small introductory generation allowance and describe the AI tools within a subscription. Check the current account terms, generation limits and permissions for the input assets before expanding a project. The wider site's model marketplace has separate purchasing and licensing arrangements, so do not treat a marketplace promotion as an AI generation entitlement.