Quick Start
Configure your first model and create your first AI agent.
Quick Start
This guide walks you through configuring your first model provider and creating your first AI agent after installing OpenAgent.
Make sure OpenAgent is running at http://localhost:14000 before continuing.
Step 1 — Add a Model Provider
Open the admin dashboard
Go to Settings → Model Providers in the sidebar.
Add your provider
Click Add Provider and choose your LLM provider. OpenAgent supports 30+ providers. For this quick start, we'll use OpenAI.
Fill in your API key:
Provider: OpenAI
API Key: sk-...your-key...Click Test Connection to verify the key works, then Save.
Select a model
After saving the provider, click Add Model and select the model you want to use (e.g., gpt-4o, claude-3-5-sonnet, gemini-2.0-flash).
Step 2 — Create a Knowledge Base (Optional)
If your agent should answer questions based on your own documents, create a knowledge base first.
Create a knowledge base
Go to Knowledge Bases → New Knowledge Base. Give it a name and choose an embedding model.
Supported embedding models include:
text-embedding-3-small/text-embedding-3-large(OpenAI)text-embedding-ada-002(OpenAI)- Local embedding models via Ollama
Upload documents
Click Upload and add your files. OpenAgent supports:
| Format | Notes |
|---|---|
| Text and scanned (OCR) | |
| Word (.docx) | Full formatting support |
| Excel (.xlsx) | Tabular data |
| CSV / TSV | Structured data |
| Plain text (.txt, .md) | Direct ingestion |
| PowerPoint (.pptx) | Slide content |
Wait for indexing
OpenAgent splits each document into chunks and generates embeddings. The status bar shows progress. Indexing time depends on document size and your embedding model.
Step 3 — Create an Agent
Open the Agents page
Go to Agents → New Agent.
Configure the agent
| Setting | Description |
|---|---|
| Name | Display name for the agent |
| Model | The LLM model this agent uses |
| System Prompt | Instructions that define the agent's behavior |
| Knowledge Base | (Optional) Documents the agent can search |
| Temperature | Response creativity (0 = deterministic, 1 = creative) |
| Max Tokens | Maximum length of each response |
Example system prompt for a customer support agent:
You are a helpful customer support assistant for Acme Corp.
Answer questions using the provided knowledge base.
If you don't know the answer, say so clearly rather than guessing.
Always be polite and professional.Test the agent
Use the built-in Chat panel to test your agent before deploying it to a channel. Type a question and verify the response quality.
Step 4 — Connect a Channel
Open Channels
Go to Channels → Add Channel.
Configure your channel
- Message @BotFather on Telegram
- Create a new bot with
/newbot - Copy the bot token
- Paste it into OpenAgent's Telegram channel config
- Click Save & Activate
Your agent is now live on Telegram. Message your bot to test it.
- Create a Slack App at api.slack.com/apps
- Enable Bot Token Scopes:
chat:write,im:read,im:history - Install the app to your workspace
- Copy the Bot User OAuth Token (
xoxb-...) - Paste it into OpenAgent's Slack channel config
- Create a Discord application at discord.com/developers
- Add a Bot to the application
- Enable Message Content Intent in the Bot settings
- Copy the bot token
- Paste it into OpenAgent's Discord channel config
- Use the OAuth2 URL to invite the bot to your server
OpenAgent includes a built-in web chat widget. After creating the channel, you'll receive an embeddable script tag:
<script
src="https://your-openagent-instance.com/widget.js"
data-agent-id="agent_xxxx"
data-channel-id="chan_xxxx"
></script>Add it to any webpage to embed the chat widget.
What's next?
You now have a working AI agent that responds to messages. Here's what to explore next:
- MCP Tool Use — Give your agent the ability to search the web, call APIs, run code, and more.
- Core Concepts: Agents — Deep dive into how agents reason and plan.
- Core Concepts: Models — Explore all supported model providers.
- Plugins — Extend OpenAgent with custom channels, tools, and providers.