Quickstart

Get up and running with Ojin in minutes.

This guide walks you through both paths, deploying a full conversational agent with the no-code Human Agent app, or driving a real-time face model from your own pipeline with the Python SDK.

Choose Your Path

Ojin offers two ways to build with real-time AI, start with whichever fits:

Deploy a Human Agent, no code (fastest)

Deploy an end-to-end conversational AI agent with a visual avatar. Ojin handles the entire pipeline, or bring your own speech-to-speech provider. No pipeline assembly required. No API key needed: you build the agent in the dashboard and the widget talks to Ojin on its own.

1. Create & configure your agent in the dashboard

2. Embed the widget on your site. One HTML tag, done

Build with the SDK. Your own pipeline

Use Ojin's real-time face models for the visual avatar layer in your own conversational pipeline. You control the full stack, STT, LLM, TTS, and the Python SDK or Pipecat turns your TTS audio into a synchronized talking avatar. Pick ojin/human-presence (flagship) or ojin/human-portrait (cost-effective) with its config_id.

1. Get your API key from the Ojin dashboard. This path calls Ojin from your own backend, so it needs one

2. Build with the Python SDK, install, authenticate, and stream your first avatar

3. Or drop the avatar into a Pipecat voice agent

Never hardcode your API key directly in your application code or commit it to version control.

Troubleshooting

Run into a snag? Check the Troubleshooting guide or reach out to support.

Next Steps

Build with the Python SDK

Install, authenticate, and stream your first avatar.

Human Agent

Learn more about the full agent product.