The Ollama CLI

Everyday Commands

ollama list      # installed models
ollama pull llama3.2
ollama rm llama3.2       # delete a model
ollama cp llama3.2 mycopy

Running and Inspecting

ollama run llama3.2
ollama show llama3.2     # parameters, size, layer info
ollama ps                # models loaded right now
ollama stop llama3.2     # unload from memory

ollama show reveals the context length a model expects, total size and quantization, useful before tuning.

Inside the Chat REPL

The interactive prompt accepts slash commands:

/show info          # details of the running model
/set temperature 0.7
/set num_ctx 4096
/exit

Help

ollama --help
ollama run --help

Key Points

• list, pull, rm and cp manage the local library.
• show and ps inspect models and memory.
• stop frees RAM used by a loaded model.
• /set tunes temperature and context in-chat.