daVenice is an AI-powered MCP (Model Context Protocol) server that connects DaVinci Resolve to Venice AI, enabling natural language control of professional video editing workflows with on-chain paymen
daVenice is an AI-powered Model Context Protocol (MCP) server that turns DaVinci Resolve into a fully voice/text-controlled video editing studio, with Venice AI as the inference backbone and MetaMask Smart Accounts powering on-chain payments for premium AI features.
What it does:
155+ MCP tools expose every DaVinci Resolve API (timeline, color, Fusion, media pool, markers, renders) to any AI assistant via natural language
- Venice AI integration: cloud transcription (Whisper), background removal, text-to-video, text-to-audio/music, text-to-image, TTS narration, chat — all invokable as Resolve tools
- Local AI fallback: Demucs-based voice/stem isolation for features Venice does not cover
- On-chain payment layer: users authorize a USDC spend budget via MetaMask Smart Accounts (ERC-7710 delegation); x402-style payment challenges settled via ERC-7715 spend permission + ERC-7710 delegation gate each paid tool call (generate_video, generate_audio, etc.)
- Solidity smart contracts (DelegationManager, Pot, enforcers) enforce spending limits on-chain
- Web chatbot UI (port 8765) for browser-based interaction with full payment gating
- Next.js landing page (daVenice) showcasing the product
Tech Stack: Python (MCP/FastMCP), Venice AI REST API, MetaMask Smart Accounts Kit, ERC-7715, ERC-7710, ERC-1271, Solidity (Foundry), TypeScript/Next.js, Tailwind CSS, Demucs, ffmpeg
Phase 1 — Core MCP Server: Built the DaVinci Resolve MCP bridge with 155+ tools covering timeline control, color grading, Fusion composition, media management, markers, rendering, and smart editing workflows.
Phase 2 — Venice AI Migration: Replaced all local AI dependencies (faster-whisper, rembg) with Venice AI cloud APIs — achieving transcription, background removal, image editing, and upscaling without GPU requirements. Added generative tools: text-to-video (seedance-2-0), text-to-audio (elevenlabs-music), text-to-image (venice-sd35), and TTS narration.
Phase 3 — On-chain Payments: Integrated MetaMask Smart Accounts (ERC-7715 spend permission + ERC-7710 delegation) to gate paid Venice AI tools (generate_video, generate_audio, generate_image) behind USDC micro-payments. Deployed Solidity contracts (DelegationManager, Pot, SpendingLimitEnforcer).
Phase 4 — Web UI & Landing Page: Launched a web chatbot (port 8765) with payment session management and a Next.js/Tailwind landing page (daVenice) demonstrating the full product experience.
Bootstrapped — built during the hackathon period. No external funding raised.
Monetization is built into the product itself: paid Venice AI tools (video generation, audio generation, image generation, TTS) are gated behind USDC micro-payments via x402-style payment challenges settled through ERC-7715 spend permission + ERC-7710 delegation. Each user authorizes a spend budget on-chain — no subscription, no signup, pay only for what you generate.
Open to grants or ecosystem funding from Venice AI, MetaMask, or ecosystem partners if the project gains traction post-hackathon.