First chat
This chapter takes about 25 minutes in the workshop.
What you will be able to do
- Send a first message to Apertus in the Ollama app or LM Studio.
- Choose between the two tools using the comparison table.
- Run five starter prompts and judge the answers.
- Recognize a confident wrong answer in the hallucination probe, a prompt that tempts the model to invent a fact.
This chapter is the first hands-on block. There is nothing new to install: the tools and the model come from Before you start. If you missed a step there, go back to Download the models first.
The two tools side by side
| Ollama app | LM Studio | |
|---|---|---|
| Chat | Yes | Yes |
| Download models | Yes, from the picker | Yes, from Discover |
| Attach files | Yes, text and PDF with + | Yes, Chat with Documents |
| System prompt | No (terminal only) | Yes |
| Temperature | No (terminal only) | Yes |
| Context setting | Yes, Settings slider | Yes, at load |
| Works offline | Yes, for chat | Yes, for chat |
| Linux desktop app | No, terminal only | Yes |
| Intel Mac | Yes, on the CPU | No |
| Used in Part 2 | Yes | No |
Use whichever tool you installed. Keep in mind that Part 2 needs Ollama.
Not sure which model size fits your machine? See Hardware.