What would you like to do?
For a first session, use a Linux machine with internet access and start with Qwen2 0.5B. The model file is about 380 MiB. Once setup finishes, ask one question and look for the response in your terminal.
Already have a task in mind? Read a recording, use a vision model to describe a picture, or train with your own examples. Check hardware and memory before downloading a larger model.
Start a conversation or connect an agent
Open Gemma or Qwen3 in the terminal · Choose a model and harness · Inspect a run