Download Ollama

Open LM Downloader, then you could see Ollama icon. Click the icon to go to the introduction page.

Click the Install Button, a separate run window opens. Ollama will start installing automatically.

If you already have Ollama installed, don't worry, this can be treated as an update to Ollama and won't affect the models you've previously downloaded.

Window users, you will see the installation prompt of Ollama itself, just follow it step by step. Mac users, you will see that the Ollama file is downloaded to the Finder and needs to be moved to the application directory yourself. The following is the window to run after installation.

Before downloading the model, set up the directory

Don't rush to download the model yet, plan your hard disk resources in advance. It is recommended that you modify the default model directory and then download the large model file. Click the settings button on the right side of Model Storage, click the Browse button, and select a directory where you want to store the model files. If you want other computers or mobile phones in the LAN to be able to use this service, you can also modify the Server Host below.

Start downloading the model

Download DeepSeek-R1

In the Ollama run window, select DeepSeek-R1, pay attention to select the parameter size you want, click the download button, and the download will start. Note that different model parameter sizes have different performance effects, and the larger the size, the greater the memory occupation, so choose the one that suits you.

Download QwQ 32B

Except for DeepSeek-R1, if your hard drive has more than 20GB of free space and your video memory is greater than or equal to 24GB. There's also the QwQ 32B model, which is comparable in performance to the DeepSeek-R1 full-blooded version.