DaVinci Resolve MCP: Automating Video Editing and Color Grading with Claude AI

The open-source davinci-resolve-mcp project (GitHub: lordhoell/davinci-resolve-mcp) has gained significant traction across the video production community by exposing DaVinci Resolve's native scripting interfaces via the Model Context Protocol (MCP), allowing AI assistants like Anthropic's Claude to directly orchestrate non-linear editing (NLE) environments.

Image source: @Hunter_Weiss / X (Blackmagic Design)

Exposing Over 440 Official Scripting APIs Through MCP

davinci-resolve-mcp wraps Blackmagic Design's official DaVinci Resolve scripting API into structured MCP tool definitions, exposing them directly to AI assistants.

The integration covers more than 440 granular tools across project management (25 endpoints), timeline editing (59 endpoints), media pool operations (27 endpoints), as well as color grading and audio track routing. Rather than relying on fragile computer vision or OS-level cursor and keyboard emulation (Computer Use), the MCP server communicates directly with DaVinci Resolve's internal scripting engine at the protocol layer.

This architecture enables an AI assistant to execute deterministic editing operations without visual inference drift:

• Timeline Manipulation: Frame-accurate cut editing, ripple deletes, and clip rearrangement across timelines.
• Metadata and Markers: Automated chapter marker creation and timecode-specific annotations derived from speech analysis.
• Audio Automation: Dynamic audio ducking that automatically attenuates background music beds under dialogue tracks.

Capitalizing on the native AI assistant and MCP connectivity introduced in the DaVinci Resolve 21.1 update, editors can trigger complex, repetitive NLE sequences entirely through structured natural language prompts.

Real-World Production Benchmarks and Target Workflows

This tool is aimed at professional video editors, post-production technical directors (TDs), and production teams seeking to reduce the manual overhead of repetitive cuts, initial grading passes, and audio cleanups.

Field testing by 15-year veteran video editor Hunter Weiss and a dedicated media test team demonstrated substantial operational time savings:

• Alexa RAW Color Grading Pipeline: Streamlined the primary grading pass for large-scale cinema camera raw files by using Claude-driven commands to assign base looks and construct node trees rapidly.
• Podcast Audio Mastering: Automated dialogue EQ adjustments, dynamic compressor settings, and LUFS loudness normalization aligned with broadcast and streaming delivery standards.
• Measurable Time Savings: Documented editorial efficiency gains of up to 10 hours per day across high-volume prep, rough assembly, and basic mastering workflows.

Studio License Requirements and Operational Considerations

Deploying davinci-resolve-mcp into an active production environment requires careful attention to specific prerequisites and architectural boundaries:

• DaVinci Resolve Studio License Required: Under Blackmagic Design's scripting licensing policy, external Python scripting APIs are exclusive to the commercial DaVinci Resolve Studio edition ($295 one-time perpetual license). The free base edition does not support external scripting calls.
• Local Filesystem Permissions and Security: The MCP server operates locally and requires broad read and write access to project databases and raw footage directories. Production environments should isolate active project directories and review permission boundaries before granting unconfined access.
• Color Grading Precision and Manual QC: While applying standardized look profiles and node setups on a per-clip basis operates reliably, maintaining exposure balance and white balance consistency across multi-camera setups still demands manual inspection and fine-tuning by a qualified colorist.

Sources

• GitHub Repository: lordhoell/davinci-resolve-mcp
• Hunter Weiss on X: @Hunter_Weiss