"author:Paolo3401" matching MCP servers:

Enables users to describe an LLM fine-tuning job once—in natural language or via YAML—and receive VRAM/FLOPs estimates plus ranked, Pareto-optimal GPU configurations (cheapest, fastest, and balanced) aggregated from live pricing across a dozen cloud providers. It integrates with Claude Code and Claude Desktop as well as a CLI and REST API, disclosing the source and freshness of every recommendation.