Structured run report

Report your test result

Report success or failure for Run Llama 3.2 3B Q4 locally with Jan on a 16GB Apple Silicon Mac. Reports are manually reviewed before they affect compatibility or recommendations; submitting this form does not change the ranking automatically.

Environment

Hardware name *Unified/System memory *

Operating system version *Comfy Desktop version

Model IDllama-3.2-3b

Recipe checkpointllama-3.2-3b-q4-k-m

Execution path

Tell us whether the Skill followed the published manual flow. Structured differences are reviewed as candidates for the next recipe revision.

How did you run this recipe? *Did the actual flow match the manual steps? *

Result

Overall result *Model loadedFirst image completedTime to first image

seconds

Peak memory, if measured

Swap behaviorOutput resolutionError or observation

LocalAIRun may publish an anonymous summary of this hardware and result after manual review. Email address, private paths, and unneeded logs will not be included.

What will be submitted

Recipe: apple-m4-16gb-llama32-3b-jan-chat · 2026.08.1

Environment: not entered · 16GB

Execution: manual · exact

Result: not selected

The report includes the fields entered above. It does not include an account, email address, device serial number, prompt, generated image, or browser history. Do not paste private paths, credentials, or raw logs into text fields.

Back to recipe