ollama

CLI

Run large language models locally (Llama, Mistral, Gemma, etc.).

Run LLMs locally without API keys — supports Llama 3, Mistral, Gemma, etc.

4 curated recipes · not yet executed by the harness#llm #local #ai #inference

Recipes

Curated from documentation and usage. Percentages are estimates, not measurements.

Pull and run a model

~96% est.

ollama run llama3

Pull a model without running

~97% est.

ollama pull mistral

Pipe text to a model

~93% est.

echo 'Summarize this text' | ollama run llama3

List installed models

~99% est.

ollama list

Homebrewbrew install ollama

Manualhttps://ollama.com/download

When to use: Local AI inference, privacy-sensitive tasks, offline LLM use
Quick examples: ollama run llama3
ollama pull mistral
echo 'Summarize this' | ollama run llama3

curl -s https://clihub.com/api/tools/ollama | jq '.agentHints.provenRecipes[] | {task, command, verified}'

Or call get_tool with slug ollama on the MCP server at https://clihub.com/mcp.