How to download models in LM Studio
LM Studio installed, internet connection
What this does
Fetches a model archive from a remote hub, stores it in the LM Studio model directory, and makes the model available in the application. After completion, the model appears in the local library ready for inference.
Steps
Open the Model Browser panel. Click the Search Models icon in the left navigation bar. The model browser shows a searchable list of available models.
Search for the desired model. Type a model name into the search bar. Filter results by size or quantization format where supported. Matching results appear in a card grid, each showing the model name, size, quantization tag, and download button.
Select quantization and download. Click the download button on the preferred model card. Choose the quantization level — smaller quantizations (such as Q4_K_M) use less disk space and RAM. The download progress bar appears at the bottom of the screen.
Confirm the model appears in the local library. Navigate to the My Models tab. The newly downloaded model should appear with a model size and last-used timestamp.
Verification
dir "$env:USERPROFILE\.lmstudio\models\*" -Recurse -Filter *.gguf | Select-Object Name
# Expected: one or more GGUF model files listed
Common failures
- Download stalls or times out — The internet connection dropped or the remote hub is rate-limiting. Pause and resume the download, or try again later.
- Insufficient disk space — A large model requires more space than available. Choose a smaller quantization variant or free additional disk space.
- Model not appearing in the library — The download did not complete cleanly. Check the download queue for a failed state and click retry.
- Search returns no results — The search query may be mis-typed or the remote API may be temporarily unreachable.