Get visual feedback on 3D models and renders

Supply the design brief and view context

Say what the object or scene is for, which element you are revising, and what a successful result should communicate. Identify the application, model or scene stage, view angle, and any reference image. For a handrail or other product form, state the intended visual relationship and the design constraints you are considering; keep exact dimensions and requirements in text.

For a learning exercise, mention the specific modeling concept you are practicing and what feedback would help you learn. For a render revision, explain which visual element changed since the prior image, such as lighting, materials, composition, or silhouette.

Choose a useful view

Share the application window or viewport containing the model using the browser's screen chooser. Select a view that reveals the feature in question, with enough surrounding geometry to show how it connects. A single perspective may hide a problem, so request a separate front, side, or top view when comparison matters. You choose each screen source; the assistant cannot open the model or switch its camera for you.

For a modeling question, show the selected object, current mode or tool, and relevant visible settings or node connections along with the viewport. Keep the cursor or selection on the feature in question and include enough surrounding geometry to show its orientation. If labels are too small, capture a separate panel view. For a render revision, compare completed frames from the same camera and ask about silhouette, composition, contrast, or material appearance. LetItSee does not transmit motion, audio, or a live viewport stream to the assistant.

Iterate on a specific observation

Ask the assistant to describe the visible form before giving a suggestion, and to identify uncertainties caused by perspective, occlusion, or lighting. You can ask whether a silhouette, connection, visible intersection, or composition changed between views, but treat these as visual observations only. Choose one change in your modeling or rendering tool, then capture the same view again for comparison. Keeping camera angle and framing consistent helps separate a real design change from a different presentation.

If the question is about a CAD object, state the intended dimensions and constraints directly. Use the CAD application's measurements and geometry tools to check them. An image may help discuss whether a shape appears balanced or whether a feature is legible, but it does not provide dependable dimensional or geometric validation.

Keep visual review and model validation distinct

A wireframe view can reveal visible edges, but a screenshot cannot reliably establish topology, constraints, units, hidden components, collision state, or exact geometry. It cannot validate fit, manufacturability, structural behavior, or compliance with a specification. Use native model inspection, measurements, simulation, and the relevant engineering review for those decisions.

For code or scene errors, share the error text, settings, project file, or logs directly if available; a screenshot can be too small to read and rarely shows the cause. Use visual capture when the rendered appearance or viewport state itself is what you want another perspective on.

Set up without starting a capture

Set up your MCP client with the full tool reference. The separate capture page can replace the active capture page for your account. It does not take a photo by itself. When your source is ready, try:

Example to adapt: Take one screenshot of the selected [Blender/CAD/render] viewport. Application and visible context: [object selected, mode/tool, and view angle]. Design goal: [goal]; I changed [feature] from [baseline]. Describe the visible silhouette or connection and any perspective/occlusion ambiguity. Suggest one visual revision to consider, but leave dimensions, topology, and geometry validation to the values and native tools I provide: [constraints or measurements].