Ollama app
The Ollama app is the quickest way to a first answer. It downloads the model for you the first time you send a message.
First message
Open the app
Start the Ollama app. A chat window opens.
Open the model picker
Click the model name in the chat window. This opens the model picker, the list where you choose a model.
Type the full model name
Type nishad/apertus:8b exactly. Use nishad/apertus:4b if your computer has limited memory (see Hardware).
apertus alone shows “No models found”. The picker looks up the name on ollama.com, so it needs the full name. Only this lookup uses the internet; chatting afterwards does not.Choose the entry with the download arrow
The entry with the download arrow is the model that is not on your machine yet. Click it.
Send a first message
Type a short message such as “Hello” and send it. The download starts.
Watch the download bar
A bar shows the download progress. The 8B model is about 5 GB, so the download takes some time on a slow connection. If you already downloaded the model as described in Download the models, there is no bar and the answer appears immediately.
Read the answer
When the download ends, the model answers in the same window. You are chatting with a model that runs on your own computer.
Interface tour
- New chat: starts a fresh conversation with an empty context, so the model does not remember the earlier chat.
- Model picker: switches between the models on your machine.
- Attach (+): adds a text file or a PDF to your message.
- Settings: a Cloud switch, a Context length slider and the Model location.
ollama run nishad/apertus:8b.To see what temperature does, read Temperature. Next, try the first prompts.