signerless/llm-checker is a JavaScript command-line tool that examines your hardware and recommends large language models (LLMs) and small language models (sLLMs) that can be run locally. The latest release recorded in the repository is v3.8.2. The provided information does not state what changed in this version, so specific changes cannot be confirmed.
This tool scores models on four criteria — quality, speed, fit, and context — to show options suited to the hardware. According to the README, it also draws on model registries from multiple sources and hardware-based memory estimation. It provides Ollama integration, and MCP integration guidance is included in the README as well. In other words, rather than merely listing models, it focuses on narrowing down candidates when you are considering local execution.
That said, the scores are comparative information meant to help you choose a model, not a guarantee of results on any specific task. The provided information does not present the detailed scoring method or actual run results. The repository's license field reads NOASSERTION, so the usage terms cannot be confirmed from this information alone. The research data momentum score is recorded as 2.3, but there is no basis for reading that as the accuracy of model recommendations or the quality of the tool. Actual suitability needs to be checked for each hardware setup and purpose.