MCP server and Blender add-on that let Claude inspect scenes, create and modify 3D objects, run Blender Python, capture viewport screenshots, and use asset or 3D-generation integrations from inside Blender.
The MCP server talks to a Blender add-on over a socket and can create, modify, delete, import, and inspect objects in the active Blender scene., The `execute_blender_code` tool can run arbitrary Python code inside Blender, so save work before use and review generated scripts before allowing execution., Optional Poly Haven, Sketchfab, Hyper3D, and Hunyuan3D workflows can download assets, import generated models, or call external services that may have their own usage limits, licenses, and terms., The README warns to run only one MCP server instance for the same Blender connection to avoid confusing the socket workflow., Do not expose the Blender socket or any remote-host setup beyond trusted local or development networks without reviewing access controls.
Privacy notes
Tool calls can expose scene names, object metadata, materials, generated Python snippets, viewport screenshots, prompt text, and asset queries to the connected MCP client and model provider., Blender MCP includes telemetry code and project terms describing collection of prompts, generated code, scene metadata, screenshots or viewport images when telemetry consent is enabled., The repository documents `DISABLE_TELEMETRY=true`, `BLENDER_MCP_DISABLE_TELEMETRY=true`, and `MCP_DISABLE_TELEMETRY=true` as environment-variable opt-outs for telemetry., The terms state that collected data may be retained, used for AI training or research, and shared in anonymized or aggregated datasets; review the current terms before using it on confidential client work., Third-party asset and generation services may receive prompts, image references, asset searches, downloads, or account identifiers depending on which integrations are enabled.
Author
Siddharth Ahuja
Submitted by
oktofeesh1
Claim status
unclaimed
Last verified
2026-06-06
Decision playbook
Review trust signals before you adopt
Signals are present but mixed. Use the checklist below to confirm the source and operational safety for your environment.
Compare context
Selected
0
Current score
63
Baseline
—
Delta
No baseline selected
No major trust-signal divergence detected in the current selection.
Source and provenance checks
Needs review
Confirm ownership and provenance before trusting install instructions.
Source link availableRequired
Open the canonical repository and verify ownership.
Done
Source provenance statusRequired
Marked as source-backed.
Done
Metadata reviewed
No reviewed flag detected in metadata.
Pending
Safety and privacy checks
Complete
Validate risk disclosures before installation or API wiring.
Safety notes presentRequired
Review the listed safety guidance before running commands.
Done
Privacy notes presentRequired
Review data handling notes before connecting accounts or secrets.
Done
Trust level risk gateRequired
Trust level does not block evaluation.
Done
Package and install checks
Needs review
Check package metadata and artifact integrity signals.
Install payload available
Install or copy payload is available for review.
Done
Package verification flag
No package verification flag provided.
Pending
Checksum metadata
No checksum provided for downloaded artifact.
Pending
Compare-driven decision checks
Needs review
Use compare context to validate trade-offs before adoption.
Compare tray has multiple entries
Add at least one more entry to compare trust differences.
5 safety and 5 privacy notes across 7 risk areas. Review closely: permissions & scopes, network access, third-party handling.
7 areas
SafetyGeneralThe MCP server talks to a Blender add-on over a socket and can create, modify, delete, import, and inspect objects in the active Blender scene.
SafetyExecution & processesThe `execute_blender_code` tool can run arbitrary Python code inside Blender, so save work before use and review generated scripts before allowing execution.
SafetyNetwork accessOptional Poly Haven, Sketchfab, Hyper3D, and Hunyuan3D workflows can download assets, import generated models, or call external services that may have their own usage limits, licenses, and terms.
SafetyExecution & processesThe README warns to run only one MCP server instance for the same Blender connection to avoid confusing the socket workflow.
SafetyPermissions & scopesDo not expose the Blender socket or any remote-host setup beyond trusted local or development networks without reviewing access controls.
PrivacyThird-party handlingTool calls can expose scene names, object metadata, materials, generated Python snippets, viewport screenshots, prompt text, and asset queries to the connected MCP client and model provider.
PrivacyTelemetryBlender MCP includes telemetry code and project terms describing collection of prompts, generated code, scene metadata, screenshots or viewport images when telemetry consent is enabled.
PrivacyTelemetryThe repository documents `DISABLE_TELEMETRY=true`, `BLENDER_MCP_DISABLE_TELEMETRY=true`, and `MCP_DISABLE_TELEMETRY=true` as environment-variable opt-outs for telemetry.
PrivacyData retentionThe terms state that collected data may be retained, used for AI training or research, and shared in anonymized or aggregated datasets; review the current terms before using it on confidential client work.
PrivacyNetwork accessThird-party asset and generation services may receive prompts, image references, asset searches, downloads, or account identifiers depending on which integrations are enabled.
Safety notes
The MCP server talks to a Blender add-on over a socket and can create, modify, delete, import, and inspect objects in the active Blender scene.
The `execute_blender_code` tool can run arbitrary Python code inside Blender, so save work before use and review generated scripts before allowing execution.
Optional Poly Haven, Sketchfab, Hyper3D, and Hunyuan3D workflows can download assets, import generated models, or call external services that may have their own usage limits, licenses, and terms.
The README warns to run only one MCP server instance for the same Blender connection to avoid confusing the socket workflow.
Do not expose the Blender socket or any remote-host setup beyond trusted local or development networks without reviewing access controls.
Privacy notes
Tool calls can expose scene names, object metadata, materials, generated Python snippets, viewport screenshots, prompt text, and asset queries to the connected MCP client and model provider.
Blender MCP includes telemetry code and project terms describing collection of prompts, generated code, scene metadata, screenshots or viewport images when telemetry consent is enabled.
The repository documents `DISABLE_TELEMETRY=true`, `BLENDER_MCP_DISABLE_TELEMETRY=true`, and `MCP_DISABLE_TELEMETRY=true` as environment-variable opt-outs for telemetry.
The terms state that collected data may be retained, used for AI training or research, and shared in anonymized or aggregated datasets; review the current terms before using it on confidential client work.
Third-party asset and generation services may receive prompts, image references, asset searches, downloads, or account identifiers depending on which integrations are enabled.
Prerequisites
Blender installed locally with permission to install and enable the repository's `addon.py` file.
uv or uvx available to run the packaged `blender-mcp` server.
Blender MCP add-on running in Blender and connected through the local socket server, usually on `localhost:9876`.
Optional API keys or account setup for third-party integrations such as Hyper3D, Sketchfab, Hunyuan3D, or asset-download workflows.
Clear project policy for whether Claude may run generated Blender Python, download assets, upload screenshots through telemetry, or use scene prompts for model-training telemetry.
Blender MCP Server connects Claude to Blender through a local MCP server and a
Blender add-on. The add-on opens a socket inside Blender, while the Python MCP
server exposes tools Claude can use to inspect the current scene, manipulate
objects and materials, run Blender Python, capture viewport screenshots, and
coordinate optional asset or 3D-generation workflows.
The project is popular because it turns Blender into an interactive creative
workspace for agent-assisted 3D work. Claude can ask for scene information,
create or modify objects, update materials, position cameras, download Poly
Haven or Sketchfab assets, and work with Hyper3D or Hunyuan3D generation when
those integrations are enabled.
These sources were reviewed on 2026-06-06. Prefer the live repository,
README, package metadata, server implementation, add-on source, telemetry code,
terms, and PyPI page for current setup steps, tool behavior, telemetry controls,
and third-party integration details.
Features
Local Blender add-on plus Python MCP server architecture.
Claude Code, Claude Desktop, Cursor, VS Code, and OpenCode configuration examples.
Scene inspection for objects, materials, cameras, and Blender state.
Object creation, editing, deletion, material assignment, and texture workflows.
Viewport screenshot capture so Claude can reason about the visible scene.
Arbitrary Blender Python execution through execute_blender_code.
Optional Poly Haven asset search and download support.
Optional Sketchfab model search, preview, and download support.
Optional Hyper3D Rodin and Hunyuan3D generation/import workflows.
Environment variables for Blender socket host and port configuration.
Telemetry opt-out controls through add-on preferences and environment variables.
Installation
Install uv first if it is not already available, then add the MCP server to
Claude Code:
Download addon.py from the repository, install it in Blender through
Preferences > Add-ons, enable the Blender MCP add-on, and click the add-on's
connect button from Blender's 3D View sidebar before using the MCP tools.
Use Cases
Ask Claude to inspect a scene and summarize objects, materials, and camera state.
Prototype simple 3D scenes, props, layouts, and material changes from text prompts.
Generate and review Blender Python before allowing it to run in the active scene.
Capture a viewport screenshot so Claude can adjust composition or object placement.
Search and import public assets or HDRIs when Poly Haven or Sketchfab workflows are enabled.
Generate rough 3D assets through Hyper3D or Hunyuan3D integrations for early concept work.
Use Claude as a scene assistant while keeping Blender as the local creative environment.
Safety and Privacy
Blender MCP is powerful because it gives Claude direct control over a local
creative application. Treat it as an execution-capable local tool, not a passive
documentation server. Review generated Python, keep backups of important
projects, and avoid running it against sensitive production assets without a
clear review step.
Telemetry deserves special attention. The README and telemetry source document
ways to disable collection, while the project terms describe possible
collection and use of prompts, generated code, scene metadata, and screenshots
or viewport images when telemetry is enabled. Disable telemetry and review the
current terms before using Blender MCP on confidential client scenes, unreleased
models, or work that cannot be used for training or shared research datasets.
Duplicate Check
No ahujasid/blender-mcp, blender-mcp, Blender socket add-on MCP, Poly Haven
Blender MCP, Sketchfab Blender MCP, Hyper3D Blender MCP, or Hunyuan3D Blender
MCP entry was found in content/mcp, content/tools, content/guides,
content/agents, or content/skills.
Show that Blender MCP Server is listed on HeyClaude. Paste this Markdown into your README — it renders the badge and links back to this page.
[](https://heyclau.de/entry/mcp/blender-mcp-server)
How it compares
Blender MCP Server side by side with 3 alternatives on trust, install, platform support, and disclosed safety notes — all from reviewed registry metadata.
3 trust signals differ across this comparison (Package trust, Source provenance, Submitter).
MCP server and Blender add-on that let Claude inspect scenes, create and modify 3D objects, run Blender Python, capture viewport screenshots, and use asset or 3D-generation integrations from inside Blender.
Official MCP server for agent-device, Callstack's device automation CLI for inspecting, controlling, debugging, recording, and collecting evidence from iOS, Android, TV, macOS, Linux, React Native, Expo, Flutter, and native apps.
✓The MCP server talks to a Blender add-on over a socket and can create, modify, delete, import, and inspect objects in the active Blender scene.
The `execute_blender_code` tool can run arbitrary Python code inside Blender, so save work before use and review generated scripts before allowing execution.
Optional Poly Haven, Sketchfab, Hyper3D, and Hunyuan3D workflows can download assets, import generated models, or call external services that may have their own usage limits, licenses, and terms.
The README warns to run only one MCP server instance for the same Blender connection to avoid confusing the socket workflow.
Do not expose the Blender socket or any remote-host setup beyond trusted local or development networks without reviewing access controls.
✓Agent Device MCP exposes structured tools backed by `AgentDeviceClient`; the docs state it does not expose generic shell execution over MCP.
Tools and CLI workflows can open apps, inspect UI, tap, type, scroll, perform gestures, wait, assert state, handle alerts, and close sessions.
Evidence workflows can capture screenshots, recordings, logs, traces, network traffic, performance samples, crash context, React profiles, and replay files.
Mutating commands should run serially against one session, and separate sessions or devices should be used for parallel work.
Prefer dedicated test devices or simulators, and require approval before entering credentials, submitting forms, changing settings, installing apps, sending messages, or touching production accounts.
✓Restrict the Airtable API key to the bases and write scopes Claude is allowed to use before enabling record or schema changes.
✓Use a least-privilege Asana token and test task or project writes in a non-critical workspace before connecting production workflows.
Privacy notes
✓Tool calls can expose scene names, object metadata, materials, generated Python snippets, viewport screenshots, prompt text, and asset queries to the connected MCP client and model provider.
Blender MCP includes telemetry code and project terms describing collection of prompts, generated code, scene metadata, screenshots or viewport images when telemetry consent is enabled.
The repository documents `DISABLE_TELEMETRY=true`, `BLENDER_MCP_DISABLE_TELEMETRY=true`, and `MCP_DISABLE_TELEMETRY=true` as environment-variable opt-outs for telemetry.
The terms state that collected data may be retained, used for AI training or research, and shared in anonymized or aggregated datasets; review the current terms before using it on confidential client work.
Third-party asset and generation services may receive prompts, image references, asset searches, downloads, or account identifiers depending on which integrations are enabled.
✓Screenshots, recordings, traces, logs, network dumps, replay files, reports, UI snapshots, typed input, and React profiles can contain private UI state, tokens, request data, customer information, or credentials.
macOS, iOS, Android, and TV automation can expose local app state, notifications, device names, package identifiers, app content, system dialogs, and permission prompts.
Network inspection artifacts may include headers, payloads, session identifiers, URLs, and API data; review before sharing or committing.
Interactive CLI runs may check npm for newer package versions unless `AGENT_DEVICE_NO_UPDATE_NOTIFIER=1` is set.
✓Airtable records, field values, attachments, and base metadata may be sent through the MCP client and model context.
✓Asana task text, project metadata, comments, assignees, due dates, and workspace details may be sent through model context.
Prerequisites
Blender installed locally with permission to install and enable the repository's `addon.py` file.
uv or uvx available to run the packaged `blender-mcp` server.
Blender MCP add-on running in Blender and connected through the local socket server, usually on `localhost:9876`.
Optional API keys or account setup for third-party integrations such as Hyper3D, Sketchfab, Hunyuan3D, or asset-download workflows.
Node.js 22 or newer and `agent-device` installed globally or project-locally.
Xcode tooling for iOS, tvOS, or macOS targets, or Android SDK and ADB for Android targets.
Device, simulator, emulator, TV, macOS, or Linux desktop target that the agent is allowed to automate.
Required local permissions such as Android device trust, iOS Developer Mode, macOS Accessibility, and Screen Recording where applicable.
Node.js 18+ installed for running npx commands and MCP server packages
Airtable Personal Access Token (PAT) from https://airtable.com/account