AI and VoIP Blog

Last updated: Feb 2025

The iTerm2 AI Ollama integration brings intelligent command suggestions, code completions, and an AI chat assistant directly into your terminal — with no data leaving your Mac. iTerm2, the popular macOS terminal emulator, ships a powerful AI plugin that connects to local models running through Ollama.

In this guide, we’ll walk you through the complete iTerm2 AI Ollama setup using the qwen2.5:7b model.

There are two AI features that iTerm introduced:
1. Command Generator: Ask AI to generate a command for a specific task as shown in the GIF above.
2. Codecierge: Ask AI about a task and it will give commands with explanations. Once you run a command AI will analyse it and give remaining commands.

First lets configure the backend to enable AI integration using the steps given below.

---

Table of Contents

1. Prerequisites
2. Step 1: Open iTerm2 Settings
3. Step 2: Access the AI Tab
4. Step 3: Install the AI Plugin
5. Step 4: Download and Run the Plugin
6. Step 5: Move the Plugin to Background
7. Step 6: Verify Available Models in Ollama
8. Step 7: Configure AI Settings
9. Command Generator
10. Codecierge
11. AI Chat
12. Conclusion

---

Prerequisites

This guide covers the complete iTerm2 AI Ollama configuration on macOS. Make sure you have the following ready:

• iTerm2 installed on your macOS. Download here.
• Ollama installed. Follow the installation instructions on the Ollama website.
• qwen2.5:7b AI model or any other model of your choice, downloaded via Ollama. Download here

Step 1: Open iTerm2 Settings

Launch iTerm2. Press command + , short keys or from the menu bar, navigate to:

iTerm2 > Settings

Step 2: Access the AI Tab

In the settings window, click on the AI tab located in the sidebar.

Step 3: Install the AI Plugin

Click on the Install the plugin button. If you have already installed the plugin the General tab will looks like below.

Step 4: Download and Run the Plugin

A download will start for the AI plugin. Once downloaded:

1. Unzip the plugin file.
2. Run the plugin installer.

Running the installer will launch a new application required for the AI features.

Step 5: Move the Plugin to Background

After installation, the plugin application may remain open. You can click on the cross button to move it to the background. The plugin will continue to run and enable AI features in iTerm2.

Step 6: Verify Available Models in Ollama

To ensure that the LLM is available, run the following command in your terminal:

ollama list

You should see an output similar to:

> ollama list
NAME            ID            SIZE    MODIFIED
smollm2:latest  cef4a1e09247  1.8 GB  4 days ago
qwen2.5:7b      845dbda0ea48  4.7 GB  10 days ago

Step 7: Configure AI Settings

Return to the AI tab in iTerm2 settings to configure the AI plugin. Click on the “Configure AI Models Manually” button and enter the following configurations:

• API: Llama
• API URL: http://127.0.0.1:11434/api/chat
• Model: qwen2.5:7b or anyother of your choise

Explanation of Settings

• API URL: This is the local endpoint where Ollama serves the AI model.
• Model: Specifies the AI model to use. In this case, qwen2.5:7b . You can go for a smaller model like smollm2:latest Download here
• Tokens: Sets the maximum number of tokens (words or word pieces) the AI model can generate.

In the AI prompt tab add the following prompt, check the below image for reference:

Interpret this prompt:
\(ai.prompt)
Generate terminal commands for macOS without explanations. Output only the command, that can be directly executed in the terminal, without any descriptions or explanations. It is Paramount that you ONLY return the command without enclosing it in a Markdown code block.

In the Features tab select the configuration you want to enable. With the settings configured, you’re ready to use the AI features.

Command Generator

Press Command + Y to launch the AI command generator UI within iTerm2. In the AI chat interface, you can now ask for command suggestions in plain English (or any supported language).
For example:

• "find top 10 files by size"
• "create a new Git branch and switch to it."

Press shift + return to submit the request. The AI will generate the appropriate command. Update the command if needed and then press shift + return to run the command.

Codecierge

To access this feature click Toolbelt → Show Toolbelt, also select the codecierge option. Now you should have a chat window on the right. You can chat with the LLM regarding any task and it will give steps to complete the task.

After each command you run the AI will analyse and update the chat window with the remaining commands.

AI Chat

To access the AI Chat feature click Session → Open AI Chat… or press shift + control + command + Y keys at the same time. You can also click the “brain” icon in the terminal as shown in the image below to access the chat.

You can chat about the terminal output in the chat like understanding the output of a command.

Conclusion: iTerm2 AI Ollama Setup

This iTerm2 AI Ollama setup gives you a fully private, free AI assistant inside your terminal. You get intelligent command construction, step-by-step task guidance via Codecierge, and a full AI chat — all powered by local models with no data leaving your Mac.

Feel free to explore different models available in Ollama and adjust the settings to suit your needs. If you want to extend your local AI setup beyond the terminal, Open WebUI gives you a full ChatGPT-style interface for the same local models, and Aider adds AI pair programming directly in your editor.

Troubleshooting

Connection Issues: Check Ollama troubleshooting steps here

Additional Resources

• iTerm2 Official Website
• Ollama Official Website
• smollm2 Model Information

---

Date: November 2024