Description
Welcome back! Today we are solving the biggest storage issue for local AI users: how to change the model download location in Ollama. Large Language Models take up massive amounts of storage. By default, Ollama saves every model directly to your primary OS drive, which can instantly max out your C drive.
In this quick video, I will show you exactly how to safely move your existing models to a secondary hard drive and force Ollama to download all future models to that new custom location.
The secret lies in setting a system environment variable named OLLAMA_MODELS.
For Windows users: Open your Start menu and search for "Environment Variables". Click "Edit the system environment variables". In the dialogue box, click "Environment Variables" at the bottom. Under the User or System variables section, click "New". For the variable name, type OLLAMA_MODELS in all capital letters. For the variable value, paste the exact folder path of your new drive location, like D:\OllamaModels. Click OK, save, and completely quit the Ollama app from your system tray. When you restart Ollama, it will now use the new drive.
For macOS users: Open your terminal and use the command launchctl setenv OLLAMA_MODELS followed by your new path. Then simply restart the Ollama application.
For Linux users: You need to edit your systemd service. Run sudo systemctl edit ollama, and add Environment="OLLAMA_MODELS=/your/new/path" directly under the Service section. Reload the daemon and restart the Ollama service.
If you already have downloaded models, do not forget to manually cut and paste your existing "blobs" and "manifests" folders from the old default directory into your newly created folder before restarting the server. This prevents you from having to re-download gigabytes of data.
If you found this quick storage management guide helpful, please hit the like button! Leave a comment down below if you ran into any path errors, and subscribe for more efficient cloud programming tricks and local AI tutorials!
Hashtags
#Ollama #ModelLocation #LocalAI #OllamaTutorial2026 #StorageManagement #MachineLearning #TechTutorial #DeveloperTools #LargeLanguageModels #PythonProgramming
Tags
change model location in ollama, move ollama models to another drive, ollama change download directory, change ollama models path, ollama environment variables, save c drive space ollama, move local llm to external drive, ollama windows setup, ollama linux custom path, macos ollama models directory, python programming tips, local ai tutorial 2026, fast ai coding tricks, developer tools tutorial, ollama blobs folder