The foundational infrastructure for spatial intelligence.
Turn images and videos into photorealistic 3D reconstructions with a single API call.
A complete API suite for spatial intelligence development
Convert images or videos into photorealistic 3D models. Powered by gaussian splatting and neural rendering.
POST /api/v1/reconstruct
{
"media_url": "image.jpg",
"quality": "high"
}Generate navigable 3D environments from text prompts. Create full scenes, not just isolated objects.
POST /api/v1/generate
{
"prompt": "modern office",
"format": "gaussian_splat"
}Modify 3D scenes with AI-powered editing tools. Change lighting, add objects, remove elements.
POST /api/v1/edit
{
"scene_id": "abc123",
"operation": "add_object"
}
Real-time reconstruction with gaussian splatting.
Process images in seconds, videos in minutes.
Standard HTTP endpoints with JSON responses.
Works with any programming language.
Official SDKs for Python, JavaScript, Unity, Unreal.
Community libraries for more platforms.
Dashboard for API usage, performance metrics.
Monitor costs and optimize integration.
API keys, OAuth2, on-premise deployment.
SOC 2 Type II certified, GDPR compliant.
Discord community, email support, SLAs.
Dedicated support for enterprise customers.
From input to 3D scene in three simple steps
Send images, videos, or text prompts via REST API. Supports all major formats: JPEG, PNG, MP4, MOV.
Our spatial AI models reconstruct or generate 3D scenes. Powered by gaussian splatting and neural rendering.
Download gaussian splats, meshes, or point clouds. Export to Unity, Unreal, Blender, or custom formats.
Pay only for what you use. No hidden fees.
Real-time 3D reconstruction using 3D gaussian primitives. Enables photorealistic rendering at 60+ FPS.
Visual Geometry Grounded Transformer for multimodal spatial understanding. Proprietary model trained on 50M+ spatial samples.
NeRF-based techniques for view synthesis and scene completion. Generate photorealistic novel views from sparse inputs.
3D-native diffusion for generative scene creation. Text-to-3D, image-to-3D, video-to-3D pipelines.
Join the waitlist for API alpha access launching Q1 2026