MLflow 3.15.0

[Daniel Lok]

MLflow 3.15.0 has been released! Here are the major new features:

• 🧩 MCP Registry: A centralized catalog for registering, versioning, and sharing Model Context Protocol servers — with semantic-versioned configs, promotable aliases, tags, auto-discovered tools, and ready-made connection instructions for Claude Code and .mcp.json. Manage it from the UI, REST API, or Python!
• 🤖 MLflow Assistant enhancements: The in-app AI assistant now supports multiple LLM providers (Claude Code, Codex, and OpenAI-compatible/Gateway endpoints) chosen from a single settings page, displays live per-session token usage and estimated cost in the composer, and is easier to set up — mlflow agent setup can enable it in one prompt, with API keys stored securely in the Gateway's LLM Connections.
• 🗂️ Sharable table views: Save named views of the Runs table — capturing columns, order, widths, filters, and sort — and share them by URL.
• ⚡ Proxy-less artifact upload/download via presigned URLs: Large artifact transfers can now bypass the tracking server and talk directly to cloud storage (e.g. S3) through presigned URLs, cutting server load and timeouts on big files. We fall back to proxied transfer automatically for backward compatibility.
• 🖼️ Multi-modal attachments in LLM judges: {{ trace }} judges created with make_judge() can now see images captured in traces — enabling evaluation of vision tasks, screenshots, and other visual content.
  
  For visual demos, check out the release post on our website! For the complete list of updates and bug fixes, see the full changelog.