PatchworkMD

LocalModelMatch · Public preview

Find a model that fits your machine.

Compare local AI models with runtime-aware memory estimates and Hugging Face discovery.

Open LocalModelMatch →

Start with your hardware

Enter and confirm your hardware before comparing models.

Compare runtime choices

Use memory estimates that account for the selected runtime and model format.

Explore available models

Browse model options and source details from Hugging Face.

Availability

Public preview

Open the preview, confirm your hardware and runtime, then compare model estimates. Check the model source and license before downloading.

Questions and answers

Are these benchmark results?

No. Fit and memory estimates are guidance, not measurements of speed, quality or stability on your machine.

Can the browser detect my hardware exactly?

Automatic detection may be incomplete. Confirm or enter your hardware manually.

Does a model fitting mean it will run?

No. Runtime compatibility, model format, context length and other workloads also matter.

Get in touch

For project questions, feedback and support.

hello@patchworkmd.dev