modl
MCP 호스팅 플랫폼
공연
RD 인기지표
RD 성장지표
비교
modl 소개
modl is a local-first tool for image generation and model fine-tuning, exposing one engine through both a command line and an MCP service. Users train LoRA adapters, generate images, upscale and remove backgrounds from the terminal, or let a coding agent do the same through tool calls, with models downloaded automatically on first use.
Written in Rust and Python, it exposes roughly 20 MCP tools covering generation, editing, training, captioning, scoring and job status. Without a local GPU, one command rents a remote GPU by the minute, artifacts sync back when a job finishes, and work can queue in the background. Model weights and job records stay in a local directory keyed by job ID, so agent reads never pass through an external service, and configs are kept as text files for reproduction. It targets developers wiring image pipelines into agent workflows, positioned between compute and models.