← Back to MCP Servers
Version ControlOfficialAdvanced
DVC MCP
Data Version Control for ML projects — track datasets and models, manage experiment pipelines, push and pull data from remote storage, and compare experiment results across versions. Essential MLOps tool for data scientists.
What is DVC MCP?
DVC MCP is a Model Context Protocol server published by iterative that lets AI assistants interact with version control services. Once configured, your assistant can call its connected tools in natural language instead of you hand-writing every request against the underlying API.
What Can You Do With It?
- Track ML datasets
- Manage experiment pipelines
- Compare experiment results
How to Use This Server
- 1. Choose your AI clientAdd DVC MCP to an MCP-compatible assistant (Claude, Cursor, VS Code, or any client listed in the official MCP docs).
- 2. Install the serverFollow the install instructions below. Most servers install with a package manager in a few minutes and need only an API key or a local credential.
- 3. Add it to your configurationWrite the server block into your client's config file exactly as the official docs publish — including any arguments, environment variables, and path fields.
- 4. Start askingRestart your client and describe the task in natural language — the assistant uses the server's tools against your real data and services.
Server Specs
- Publisher
- iterative
- Category
- Version Control
- Difficulty
- Advanced
- Tags
- DVCData VersioningMLOpsPipelines
Get It Set Up
Security note: MCP servers run with the credentials you give them. Only install servers you trust, review tool lists before enabling them, and use scoped credentials (read-only tokens where possible). Axonix is an independent directory, not the publisher of this server.