MarvinLabs vram_fit

Overview

MarvinLabs vram_fit is a Model Context Protocol server that answers whether a QLoRA fine-tune of gemma-3-4b-it will fit on your GPU. It provides verdicts based on measured probes from both AMD and NVIDIA hardware, helping you plan training runs without guesswork.

The server runs as a remote endpoint at https://marvinlabs.org/mcp with no authentication required and nothing to install locally. Current release is v0.1.0. Visit https://marvinlabs.org/vram for details and to learn more about the project.

MarvinLabs vram_fit is listed in the Development category and works with MCP clients including Claude Desktop, Claude Code, Cursor, VS Code, Windsurf, Zed, and others. If you maintain this server, claim your listing to earn the Verified badge and keep information up to date.

Installation

Settings → Connectors → Add custom connector
Name: MarvinLabs vram_fit
URL:  https://marvinlabs.org/mcp