Show the full silhouette
Avoid cutting off the top, base, or projecting parts of the object. If the source photo crops a handle or tail, Tripo must guess how it continues outside the frame.
Turn one clear picture of an object into a downloadable 3D model. Upload a product photo, prop, or character reference, choose texture detail, then inspect the generated GLB from every angle.
Example
Example input: a single product on a clean background
The lamp image below is a sample input created for this page—not a Tripo-generated result. Its full silhouette and uncluttered background make the object easier to identify. Hidden surfaces still have to be inferred by the model.
A white or plain background can help separate the subject from scenery, but it cannot supply missing back, underside, or internal details. Review the finished mesh rather than assuming the photo was reconstructed exactly.
Input guidance
Prepare a picture for 3D reconstruction
The source image is the only visual evidence the model receives. A little preparation can make the result easier to inspect, though it cannot guarantee exact geometry.
Separate object and background
A neutral or contrasting background makes it easier to see the subject's outline. Remove unrelated objects when possible; a busy scene may be interpreted as part of the model.
Expect inferred hidden surfaces
The back, underside, and occluded parts cannot be recovered from a single view. Rotate the GLB after generation and compare those areas against any additional references you have.
Check before downstream use
A generated mesh can be useful for concepting and prototyping, but inspect texture placement, dimensions, and topology in your target 3D software before a final game, print, or product workflow.
Workflow
How image to 3D works
1. Choose one reference
Upload a clear photo or illustration of one complete object. We resize it for the request and review it for safety before sending it to Tripo P2.
2. Choose the texture
Use geometry-only for your own materials, or choose standard, detailed, or extreme textures. The exact credit charge appears before generation.
3. Inspect the GLB
Rotate the saved model in your browser, then download the GLB. Look closely at the back and hidden edges, because one photo cannot show every side.
Credit pricing
One model, one credit price
Image to 3D and Text to 3D use the same credit prices. Choose the texture setting that fits your project; the exact credit charge is shown before generation.
| Texture | Credits per model |
|---|---|
| No texture | 15 credits |
| Standard | 16 credits |
| Detailed | 17 credits |
| Extreme | 18 credits |
Failed requests refund reserved credits. See credit packs and plans before generating.
Choose by output
AI camera angle or 3D model: which should you use?
Use AI Camera Angle when you need another 2D view of a photo. Use Image to 3D when you need a rotatable GLB from a picture. Use Text to 3D when you need a GLB from a written object description.
| Goal and tool | Input | Output | Best for | Important limit |
|---|---|---|---|---|
| Create another camera view of an existing photoAI Camera Angle | One image | A new 2D PNG | Front, side, back, top-down, low-angle, and close-up images | No mesh, GLB, measurements, or exact hidden geometry |
| Create a rotatable 3D model from a pictureImage to 3D (current page) | One object image | A downloadable GLB 3D model | Concept models, prototypes, and assets you need to rotate | Hidden surfaces and scale are inferred, not measured |
| Create a new 3D object from a descriptionText to 3D | One object prompt | A downloadable GLB 3D model | New props, characters, and object concepts without a source photo | Not an exact reconstruction or automatically ready for 3D printing |
All three workflows generate new content. Review labels, textures, proportions, and unseen surfaces before publishing or using an output in production.
Related workflows
Choose the right type of output
Frequently asked questions
Can one picture recover every side exactly?
No. A single image only shows one view. Tripo P2 infers hidden surfaces, so inspect the GLB from all sides and revise your input if exact product geometry matters.
What image makes a useful input?
Use one fully visible object with clear edges and limited background clutter. Tiny, cropped, or heavily occluded subjects leave more shape to inference.
Is this the same as AI Camera Angle?
No. Camera Angle creates new 2D pictures. This tool creates a 3D model file that you can rotate and download.
Will I receive a GLB file?
Yes, this workflow saves and delivers a GLB model. It does not advertise FBX or production-ready measurements.