Describe a model, inspect it in a live 3D viewport, then ask for the next change. Mixos AI works with editable geometry, separate objects and materials, so your first prompt can become a project you keep refining.
Mixos AI is available in this build. Conversational tools require paid access or purchased credits.
A CREATIVE STARTING POINTStart with a recognizable shape and a few concrete requirements. The assistant uses the scene's modeling tools, and the viewport lets you check proportions and placement before asking for more detail.
Name the object, its proportions and the parts you need to edit separately. A compact speaker with a rounded enclosure and an inset driver gives the assistant a clearer starting point than a style word alone.
Make the enclosure wider, move the grille or switch the surface to brushed copper. Review the actual objects in the viewport, then continue the conversation or undo a turn.
Save the scene, continue editing by hand in Studio, or compose a render. Explicit automatic mesh generation is another route when you want a textured asset from text or a reference image.
The viewport makes the result visible. Inspect the shape, placement and materials before moving on to the next change.
Keep the robot, workshop and props editable in 3D. Use that view as the starting point for an image with richer surfaces, lighting and atmosphere.
Real app screenshot
Illustrative AI finishThis comparison illustrates the scene-to-image workflow. The finished image was created for this page with ImageGen; it is not a capture of an in-app generation. AI interpretation can change details.
Ordinary modeling requests use the modeling tools in the open scene. This suits product forms, blockouts and objects whose dimensions, components or placement you want to keep changing. Ask for separate parts where you need independent control.
When you explicitly request automatic generation, a paid model engine can create an asset from text or an image. Inspect the resulting geometry and materials before relying on it. Generated topology and fine details vary; a generated asset is not a guarantee of a clean production mesh.
A model is only one part of a useful scene. Change its materials, add lights, set a camera and place it beside other objects. You can move between conversational changes and manual Studio edits as the project develops.
“Build a compact speaker with a rounded olive enclosure, copper driver and dark grille. Keep the enclosure, driver and grille as separate objects.”Open Mixos AI
“Make the enclosure wider. Keep the driver centered, move the grille slightly forward and give the driver a brushed copper finish.”Open Mixos AI
Example briefs to adapt to your project. Concept images show a creative direction; live results depend on your prompt and the tools used.
Build, finish and animate real 3D scenes. Choose an example, play the walkthrough or jump to any step.
“Build a compact studio speaker with an olive enclosure, a copper driver and a fine metal-mesh grille. Keep the components separate.”
The chat explains the change. The viewport shows the result. Play all steps, or use the numbered controls to explore.
Prepared walkthroughs with editable 3D, animation and illustrative AI images. Prompts and replies are scripted; playback uses no credits. AI video generation is a separate paid step, illustrated above.
Mixos AI is available in this build. Open the workspace with your Mixos account; conversational tools require paid access or unspent purchased credits. The expanded creation workspace is rolling out in preview builds.
AI 3D modeling uses an assistant or a generative model to help create and change 3D geometry. In Mixos AI, conversational modeling operates on the editable scene. Explicit text-to-3D or image-to-3D generation creates a mesh through a separate paid model engine.
Conversational modeling edits objects and geometry in the current scene, so you can ask for changes to individual parts. Text-to-3D generation produces an asset from a prompt. Both can be useful: build for control over components, or generate a starting mesh and inspect it before further work.
Yes. Scene objects remain editable, and you can continue working in Studio. The kind of edits available depends on the object: separate modeled components can be adjusted independently, while a generated mesh may require mesh editing or cleanup.
Mixos is a creative 3D workspace. Use it for visual models, scene building and product concepts. Do not assume a generated mesh has engineering tolerances, validated dimensions or fabrication-ready topology; inspect the output in the tools required by your project.
No. The Mixos AI preview combines chat and a live viewport in the browser. You can use the released Studio app for manual modeling today. External Blender or MCP setup is optional, not required for this workflow.
Editing in the released Mixos apps is free with a Mixos account. AI generation consumes credits. Production exports require a paid plan or licence; the free plan includes a 512px watermarked render preview. See Pricing for current allowances. Watching the demonstrations on this page does not spend AI credits.
See Pricing for current AI credits and export allowances. Studio is available now for manual 3D creation.