Customize your input with more control.
Customize your input with more control.
Hint: Drag and drop image files from your computer, images from web pages, paste from clipboard (Ctrl/Cmd+V), or provide a URL. Accepted file types: jpg, jpeg, png, webp, gif, avif, heic, heif
Type # to reference inputs.
Customize your input with more control.
Your request will cost $0.005 per unit.
Your request will cost $0.005 per unit.
Meta's SAM 3D delivers state-of-the-art 3D reconstruction from single images at $0.005 per generation. Trading traditional multi-view capture workflows for instant single-image processing, it generates detailed 3D meshes and human body models in under a second. Built for developers shipping AR/VR experiences, e-commerce product visualization, and character animation pipelines where speed and cost efficiency matter.
Use Cases: Product Visualization | Character Rigging | AR/VR Asset Creation
At $0.005 per unit, SAM 3D operates 8-30x more cost-effectively than alternatives while maintaining production-ready quality for both object reconstruction and human body estimation.
| Metric | Result | Context |
|---|---|---|
| Inference Speed | ~0.5 seconds | Comparable to TripoSR, faster than multi-view alternatives |
| Cost per Generation | $0.005 | 200 generations per $1.00 on fal |
| Input Requirements | Single image | Eliminates need for multi-view capture or depth sensors |
| Related Endpoints | SAM 3D Objects, SAM 3D Body, SAM 3 Image | Objects vs Body vs Segmentation variants for different reconstruction needs |
SAM 3D extends Meta's Segment Anything foundation with unified 3D reconstruction capabilities. Where traditional photogrammetry demands dozens of images and minutes of processing, SAM 3D generates textured meshes from a single input through learned priors about object geometry and human anatomy.
What this means for you:
Instant Asset Generation: Single-image input eliminates complex capture rigs, upload a product photo and receive animation-ready 3D models in under a second for rapid prototyping workflows
Dual Specialization: Separate SAM 3D Objects and SAM 3D Body models optimize for either general object reconstruction or human body/shape estimation with SMPL-compatible output
Production-Ready Output: Generates textured meshes compatible with standard 3D pipelines, eliminating manual cleanup for Unity/Unreal Engine integration
Cost-Efficient Scaling: At $0.005 per generation versus $0.04-0.15+ for alternatives, process thousands of assets for e-commerce catalogs or game development at fraction of traditional costs
| Spec | Details |
|---|---|
| Architecture | SAM 3D |
| Input Formats | Single RGB image (JPEG, PNG, WebP) |
| Output Formats | 3D meshes (textured), SMPL body models (SAM 3D Body variant) |
| Reconstruction Type | Object geometry + Human body/shape estimation |
| License | Commercial use enabled |
API Documentation | Quickstart Guide | Enterprise Pricing
SAM 3D Objects – The SAM 3D RLE endpoint prioritizes segmentation workflows with Run Length Encoding output for mask-based applications at identical cost. SAM 3D Objects specializes in full 3D object reconstruction with textured mesh output for asset generation pipelines.
SAM 3 Image – SAM 3D's reconstruction endpoint trades 2D segmentation for full 3D geometry generation at the same per-inference cost. The image-to-image variant remains ideal for mask generation, video frame segmentation, and annotation workflows where 2D outputs suffice.
SAM 3D Body – SAM 3D emphasizes human body reconstruction through its specialized Body variant with SMPL compatibility for character animation. This endpoint prioritizes anatomically accurate human models for gaming, virtual try-on, and motion capture applications at identical pricing.
Tripo3D Image to 3D – SAM 3D delivers comparable sub-second reconstruction speeds while offering dual specialization through Objects and Body variants. Tripo3D prioritizes general object reconstruction with emphasis on texture quality and geometric detail for product visualization at competitive speeds.