scenario-labs/blender-plugin

Scenario for Blender – AI‑powered content generation inside Blender

What it is – A Blender add‑on (Python extension) that lets you call Scenario’s cloud AI services directly from the 3D‑authoring environment. It can generate images, videos, 3D meshes, PBR materials, audio, 360° environments and even whole block‑out layouts from natural‑language prompts or reference assets, then import the results back into the current Blender scene.

Key capabilities

• Image & video generation – Turn a text prompt or reference images into textures, background plates, or full‑length video clips.
• 3‑D mesh creation & editing – Ask the AI to create new geometry or apply edits to a selected mesh, then pull the edited mesh into Blender.
• Material synthesis – Produce complete PBR texture sets and automatically assign them to a material slot.
• Audio synthesis – Generate speech, music or sound‑effects and place them on the video sequencer.
• Render‑assisted generation – Use a scene capture (still or video) as a visual cue for the AI to produce matching assets.
• Blockout tool – Provide a short scene description and receive a grey‑box layout that can be refined further.
• Environment creation – Generate HDRI panoramas or up‑scaled 360° back‑drops and apply them to the World.
• Studio view & floating composer – UI panels that let you run the above “lanes” without leaving the viewport.
• MCP agents – A local scenario‑blender server can expose the scene to external agents (e.g., Claude Code, Cursor) that can read the scene and request paid AI work on your behalf. The server runs on http://127.0.0.1:9876/mcp and can also be used head‑lessly via blender --background … --command scenario_blender.

How to get it

1. Enable online access in Blender (Edit → Preferences → System → Network).
2. Add the remote repository https://blender.scenario.com/repo/index.json via Edit → Preferences → Get Extensions → Add Remote Repository.
3. Install “Scenario” from the list; the add‑on appears in Preferences → Add‑ons.
4. Paste your Scenario API key/secret (obtained from the Scenario web console) into the add‑on’s credential panel and test the connection.
5. Choose an output folder; generated assets will be saved there and also appear in the Generations panel.

(You can also download a ZIP from the GitHub releases page and install it manually, but that copy won’t auto‑update.)

Typical workflow

1. Open the Scenario sidebar (N in the viewport).
2. Select a lane (Image, 3D, Audio, etc.), write a prompt, attach any reference files, and review the cost estimate (Scenario charges Creative Units).
3. Click Generate – the add‑on sends the request to api.cloud.scenario.com, receives the result, and stores it locally.
4. Use the provided actions to import the asset (e.g., apply a generated texture, add a mesh, place audio on the sequencer).
5. Iterate or combine multiple lanes to build a full scene.

Limitations & caveats

• Requires a paid Scenario account; each generation consumes credits (Creative Units).
• All prompts, parameters and reference files are sent to Scenario’s cloud service before generation; privacy details are in the repo’s PRIVACY.md.
• The add‑on is marked Experimental – some lanes may be unavailable or have reduced functionality on certain Blender versions.
• Agents (MCP) can execute arbitrary Python if you opt‑in, which is a security risk; it is disabled by default.
• Offline ZIP installations do not receive automatic updates; you must reinstall to get new features or bug fixes.

License – The extension code is released under GPL‑3.0‑or‑later (required for Blender add‑ons). Bundled SDK wheels retain their original licences; see LICENSE, SDK_BUNDLE.md and TRADEMARKS.md for details.

Where to learn more

• User guide: docs/USER_GUIDE.md (covers each lane, cost preview, output handling).
• API docs: https://docs.scenario.com.
• MCP protocol reference: docs/MCP.md.
• Support: open issues via the repository’s Support page or email support@scenario.com (do not post credentials publicly).

---

Scenario for Blender brings generative AI directly into the 3‑D creation loop, letting artists prototype assets, environments and even whole scenes without leaving Blender.