Turn a photo of a part into an STL file for 3D printing with a single Gemini prompt
Upload a photo of an object with approximate dimensions to Gemini 4 Argon - get an interactive 3D studio with sliders and STL export, ready for your slicer
- 1.Open the Gemini app (gemini.google.com) and make sure the Gemini 4 Argon model is selected. If the model is not in the list - it has not reached your account yet; user polls show the rollout is gradual.
The Canvas feature and the Argon model are available on paid Google AI Pro / Ultra plans. The free Gemini Argon plan does not include them.
Gemini ↗About this tool - 2.Take a clear photo of the part or object you want to print. Place the object on a flat background, shoot from above or the front so that key geometric features are visible. Use a ruler or marker to note approximate dimensions right on the photo - or write them down separately for the prompt text.
The more even the lighting and the more contrast between the object and the background, the more accurately the model will detect the shape.
- 3.In the Gemini input box, attach the photo (via the «+» icon or by dragging) and paste the following prompt, replacing the text in square brackets with your values.
Canvas will open in the response - it is an interactive widget with a 3D preview and sliders. If a plain text response appears instead of Canvas, repeat the prompt, explicitly adding the phrase «open in Canvas».
Gemini ↗About this toolPromptOpen Canvas. I'm uploading a photo of a [bracket / furniture foot / decorative clip - describe the object in one or two words]. Approximate dimensions: width [X] mm, depth [Y] mm, height [Z] mm. Build an interactive 3D studio where I can adjust size and style with live sliders, then export the result as an STL file suitable for 3D printing.
What this prompt doesThe English prompt asks Gemini to open Canvas, build an interactive 3D studio with size and style sliders, and prepare an STL export. Replace the text in square brackets: briefly describe the object and enter approximate dimensions in millimeters. - 4.When Canvas opens, adjust the shape and dimensions with the sliders right in the widget. Check that the dimensions match the real ones - use the size grid in the scene as a reference.
The interactive visualization is a way to refine the geometry, not the final file. The «Copy» button saves a PNG snapshot, not a 3D mesh.
Gemini Canvas ↗About this tool - 5.Request an STL export directly in the chat, following the prompts inside Canvas. If the export button does not appear, ask Gemini to write a Python script that generates an STL file from primitives based on your dimensions - run the script locally and get the finished file.
Always open the generated mesh in a slicer (Cura, PrusaSlicer, OrcaSlicer) and check for holes, thin walls, and non-manifold surfaces before printing. STL does not store dimensions in millimeters - scale is set in the slicer.
Gemini ↗About this toolPromptExport the current Canvas 3D model as a downloadable STL file. If direct export is not available, generate a Python script that writes an STL mesh matching the current dimensions.
What this prompt doesThe English prompt asks Gemini to either provide the STL directly from Canvas or generate a Python script that writes an STL file with the current dimensions.
- subscriptionGoogle AI Pro (Gemini 4 Argon)The Argon model and Canvas are available on the paid Google AI Pro plan; the weekly quota can be consumed quickly - users report that 10 - 15 large prompts use up to 60% of the limit.
- free3D printing slicer (Cura / PrusaSlicer / OrcaSlicer)Free open-source programs for slicing STL into G-code; installed locally on any computer.
Why today
AI services change fast - interfaces and free limits may differ from what's described.