Local Compute Lab LCL is a bilingual, source-backed workbench that matches open-weight model requirements to NVIDIA, AMD, and Apple devices. Compatibility states and market observations are dated and carry their sources; recommendations are computed in the browser and require no account. The lab covers:

Model-to-device matching across NVIDIA, AMD, and Apple with memory assumptions and compatibility evidence

Dated market observations, concept cards, and benchmark notes

Comparison of local and cloud deployment assumptions, plus a change log and the stated methodology

The full workbench needs JavaScript. With JavaScript disabled, this page lists what the lab covers; device matching, benchmarks, and scenario comparison are unavailable.