Quick Start#
Get from zero to your first local-AI API call in under 5 minutes.
Prerequisites
You need Google Chrome (or a Chromium browser) and a machine with at least 8 GB RAM to run a model locally.
Install Ollama
Download and install Ollama for your operating system. Ollama runs a local inference server at http://localhost:11434.
# macOS (Homebrew)
brew install ollama
# Linux
curl -fsSL https://ollama.com/install.sh | sh
# Windows — download the installer from https://ollama.com/downloadVerify Ollama is running:
curl http://localhost:11434
# → Ollama is runningPull a model
Pull a model from the Ollama library. llama3.2 is a good starting point — small enough to run on most machines, capable enough for most tasks.
ollama pull llama3.2Wait for the download to complete, then verify:
ollama list
# NAME ID SIZE MODIFIED
# llama3.2:latest a80c4f17acd5 2.0 GB just nowInstall the Chrome Extension
Install the PullO extension from the Chrome Web Store:
- Go to runtime.co/extension or the Chrome Web Store
- Click Add to Chrome
- Confirm the permissions prompt
The PullO icon will appear in your browser toolbar.
Connect to PullO
- Click the PullO icon in your Chrome toolbar
- Sign in with your Runtime.co account (create a free account at runtime.co if you don't have one)
- Click Connect in the extension popup
- The status indicator will turn green — your local Ollama instance is now reachable through PullO
PullO Extension
┌────────────────────────────┐
│ ● Connected │
│ │
│ Models online: 1 │
│ llama3.2:latest │
│ │
│ Queue depth: 0 │
└────────────────────────────┘Make your first API call
Your models are now accessible at the PullO API endpoint. Grab your API key from the dashboard, then run:
curl https://api.runtime.co/v1/chat/completions \
-H "Content-Type: application/json" \
-H "Authorization: Bearer sk-local-YOUR_KEY" \
-d '{
"model": "llama3.2",
"messages": [
{
"role": "user",
"content": "Hello! What is PullO?"
}
]
}'Expected response
You should receive a response like this within a few seconds:
{
"id": "chatcmpl-abc123",
"object": "chat.completion",
"created": 1720000000,
"model": "llama3.2",
"choices": [
{
"index": 0,
"message": {
"role": "assistant",
"content": "PullO is a pull-based tunnel that lets you expose your local Ollama models as OpenAI-compatible APIs without port forwarding or ngrok."
},
"finish_reason": "stop"
}
],
"usage": {
"prompt_tokens": 14,
"completion_tokens": 37,
"total_tokens": 51
}
}What's next?#
- Read the API Reference for full endpoint documentation
- Set up CLI tools and coding agents to use your local models with Claude Code, Cursor, or Continue.dev
- Invite teammates to share access to your models