Inputs
References, images, prompts, masks, palettes and spatial views enter only where the receiving tool can use them.
Knowledge base
Complex work.
Clear operation.
Learn the system through creative tasks, not technical graph theory.
First session
From install to first workflow.
- 01
Activate Studio
Once invited, verify your account, obtain pilot access and install Studio from your account. The online trial starts on first activation, not account creation.
- 02
Choose execution
Add a supported provider API key in Settings, or install Brain on the same compatible workstation for local jobs.
- 03
Bring a source
Drag an image, paste a screen capture, import an existing 3D reference or begin from a prompt.
- 04
Connect the path
Route source, direction and generator into a visible workflow. Check references, settings and estimated cost before generating.
- 05
Save and check
Save the project and reopen it to check the result. During the pre-release pilot, keep independent backups and export important outputs separately, especially when moving work between canvases or computers.
Core concepts
Five ideas unlock the canvas.
Context
Global Context carries shared direction; Subject Context keeps subject identity separate when the workflow needs it.
Chain
Chain Context can pass relevant creative history forward. Use it when continuity matters and avoid it when a clean restart is better.
Branches
Duplicate a path to compare alternatives without destroying the direction that already works.
Templates
Save repeated generator configurations as reusable personal tools.
Local AI
Brain manages the machinery.
Studio should stay focused on creative direction. Brain owns compatible local models, LoRAs, GPU jobs and diagnostics.
Same workstation
Install Studio and Brain on a compatible Windows workstation. Connect the local endpoint and check the available models.
Your choice
Keep Brain uninstalled if you only use cloud providers. There is no need to set up local models to explore Studio.
Model setup
Download or import a supported model inside Brain after reviewing and accepting its separate licence.
Brain never silently falls back to a cloud API. A local job either runs on the selected Brain or reports an actionable local error.
System requirements
Choose the path your hardware supports.
VEINDRA StudioWindows 10 or 11, x64
Modern multi-core CPU, 16 GB RAM recommended, current GPU drivers and sufficient project storage.
Cloud workflowsInternet + provider access
A supported Google or OpenAI API key. Provider availability, pricing, data policies and model terms apply.
Brain / FLUX localNVIDIA GPU, 24 GB VRAM recommended
The tested reference workstation uses an RTX 3090. Requirements vary with model and resolution; lower-memory paths are not promised unless validated.
Network deploymentNot part of the current supported offer
Use Studio and Brain on the same workstation. Do not expose Brain to a public or shared network. Corporate and remote deployments need separate validation.
Support
When something breaks, keep the evidence.
Describe the workflow, provider or local model, exact error and the action that triggered it. Never send project assets or API keys unless a specific secure support process requests them.
Next / put it to work