Scan a room with your phone, then talk to it from Claude, ChatGPT, Gemini, or Copilot. Free forever for personal use. Pro Beta free while we’re in beta.
Backed by 5 granted US patents (priority July 2012) · Built by an NVIDIA Inception member
Forever · no credit card
Free during beta · will be $14.99/mo at GA
Free today. We’ll email you at least 30 days before any charge begins.
Tailored to your deployment
Pro Beta is free while the product is in beta. We’ll notify you at least 30 days before we start charging. Cancel any time from your account.
Even on Free, your capture is a real scene an LLM can read. MCP isn’t locked behind a paywall.
Connect Claude, ChatGPT, Gemini, or Copilot to your captures over the Model Context Protocol. They can read the scene; they can’t silently rewrite your room.
Capture from any iPhone or Android via the browser PWA. We reconstruct the scene server-side; you get back an explorable splat with a shareable viewer link.
5 granted US patents with a July 2012 priority date, covering the spatial runtime underneath every capture.
Raku Capture is built by an NVIDIA Inception member. We’re bringing NVIDIA-accelerated 3D Gaussian splatting (the Brush trainer on A10G / L40S) to reconstruction, so a phone scan turns into an explorable splat in minutes — coming via the program’s NGC catalog and partner cloud credits. Read the announcement →
RakuAI is built on AWS and an AWS Activate participant. Production reconstruction runs on Azure today; AWS GPU capacity (G5 / G6e) is on our roadmap, honestly gated until quota lands.
“AWS” and “AWS Activate” are trademarks of Amazon.com, Inc. or its affiliates. RakuAI’s participation does not imply that AWS endorses RakuAI’s products.
Yes. Pro Beta is free while Raku Capture is in beta. We’ll email you at least 30 days before any charge begins, and you can stay on Free if you don’t want to upgrade. The expected price at general availability is $14.99/month.
One uploaded scan that we reconstruct into a 3D scene. The day counter resets at midnight UTC. Re-opening or sharing an existing capture doesn’t count.
On Free, captures and their reconstructed scenes are deleted 7 days after upload. On Pro Beta it’s 90 days. You can always re-upload the source video. Enterprise plans get custom retention.
Yes. The MCP runtime is production-deployed and available on every tier — Free included. Connect Claude, ChatGPT, Gemini, or Copilot to your captures and they can read the scene over the Model Context Protocol.
Server-side splat training on NVIDIA A10G / L40S is on the roadmap via NVIDIA Inception cloud credits. Today reconstruction runs on CPU; once GPU lands we’ll route Pro Beta and Enterprise traffic to it first.
That’s Enterprise. Email partners@rakuai.com with your use case and rough scale and we’ll get back to you.
It runs in iPhone Safari and Android Chrome — no install. Scan a room, watch it reconstruct, then point Claude at it over MCP.