Ollama Cli Chat

For Offline Developers

A lightweight CLI-First alternative to WebUIs for interacting with Ollama, designed to simplify local and offline LLM usage via Docker or direct API access.

Perfect if you want:

• No browser
• No cloud dependency
• Fast terminal-based interaction with your local LLMs

---

Features

• 🧠 Chat with Ollama models directly from your terminal
• 📴 Built for Offline-First Developers
• 🐳 Works seamlessly with Dockerized Ollama(direct API call)
• 🔌 Uses Ollama’s native API (0.0.0.0:11434)
• 🌐 Supports local-network Ollama instances via OLLAMA_BASE_URL
• ⚡ Minimal binary, zero runtime dependencies

---

Installation

1. Download the Binary

Download the correct binary for your system:

• MacOS (Apple Silicon): arm64
• Linux: amd64
• Windows: (Clone the project and build it for your target architecture.)

Download the binary

Download Binary (Linux amd64)
Download Binary (Mac arm64)

---

2. Move Binary to ~/bin(create if it doesn’t exist)

mkdir -p ~/bin
mv ollama-cli-chat-("set arch name") ol
mv ol ~/bin/
chmod +x ~/bin/ol

You may rename ol to anything you prefer.

Ensure ~/bin is in your $PATH:

export PATH="$HOME/bin:$PATH"

MacOS

Apple's Gatekeeper blocks apps from unidentified developers by default.
This is a security feature, not necessarily an indication of malware.

Steps to Run It:

• Go to System Settings → Privacy & Security
• Scroll down to Security section
• Look for a message like: "ollama-cli-chat-arm64" was blocked... or "ol" was blocked... (if you renamed it)
• Click "Allow Anyway"
• Confirm with your password/Touch ID

---

Configure Ollama Base URL (Local Network)

To use a running Ollama instance exposed within your local network, set the OLLAMA_BASE_URL environment variable.

Add the following line to your shell config:

• ~/.bashrc
• ~/.zshrc
• or any other shell profile you use

export OLLAMA_BASE_URL=http://<NETWORK_IP>:<PORT>

Apply the changes:

source ~/.bashrc
# or
source ~/.zshrc

This allows the CLI to communicate with Ollama running on another machine or container in the same network.

---

Running Ollama with Docker

If Ollama is not already running, use Docker.

docker-compose.yml

services:
  ollama:
    image: ollama/ollama
    container_name: ollama
    ports:
      - "11434:11434"
    volumes:
      - ./ollama-v:/root/.ollama
    restart: unless-stopped

Start Ollama:

docker compose up -d

---

Usage

Start the CLI:

ol

---

Help

ol help

Multiline Input

• Start with .. and press Enter
• End with .. and press Enter

---

Available Commands

Command	Description
\h	Show Help
\q	Quit the session
\i	Show chat information
\c	Show the currently selected model
\m	Select a different model
\p	Pull/download an available LLM model
\d	Delete an existing model
\n	Start a new session (clears last Q/A)

---

License

MIT