BlenderGPT ships in two parts: a web app that does the heavy lifting in the cloud, and a bridge add-on (currently v2.2.0, about 9 KB) that connects it to Blender 4.2 or newer. Both downloads are free.
The BlenderGPT app runs in your browser: type prompts, drop reference images, browse the asset library, and manage your plan. Nothing to install, and the free tier needs no payment details.
The BlenderGPT addon is a small bridge extension. It holds the two-way sync: generated models drop straight into your open scene, and conversational edits apply live inside Blender.
The whole setup takes about 2 minutes on a fresh Blender 4.2 install. If anything fails along the way, the troubleshooting guide covers the common causes.
Sign in to the BlenderGPT app and download the add-on zip from the official download endpoint (the file is named blendergpt_addon_v2.2.0.zip). Keep it zipped.
In Blender, open Edit → Preferences → Add-ons, choose Install from Disk, and select the zip. Blender 4.2+ handles the extension format automatically.
Tick the checkbox next to BlenderGPT in the add-on list, then link it to your account so the app can see your open scene.
Ask the assistant to import any free asset. If it appears in your viewport, two-way sync is live and you are set up. If nothing arrives within a minute, the usual suspects are a declined network permission or a firewall between Blender and the service — both take seconds to check and are covered in the troubleshooting guide.
Generation happens in the cloud, so the requirements are the ones Blender itself sets — there is no extra GPU or VRAM demand from BlenderGPT.
Blender
Version 4.2.0 or newer (the add-on manifest sets this minimum)
Operating system
Windows, macOS, or Linux — anywhere Blender 4.2 runs
Network
An internet connection for the cloud generation and asset library
Free plan limits and paid allowances are listed on the pricing page; version history lives in the changelog.