Getting Started
Create a Botyard organization, add an LLM provider credential, and deploy your first bot. A step-by-step setup guide with pointers to what to configure next.
This guide takes you from a new Botyard organization to a running bot with an LLM provider credential attached.
1. Create your organization
Sign in to app.botyard.io. If this is your first time using Botyard, create an organization when prompted. Your organization owns the bots, credentials, MCP servers, and API keys you create.
Directions in these docs that start at User menu begin with your name in the bottom-left corner of the app: click it to open the menu.
2. Add provider credentials
Bots need a model provider before they can answer with an LLM. You can either bring your own provider API key or use an account-backed provider when available. If you plan to run your bot on Botyard credits (see step 4 of section 3), you can skip this section.
- Open User menu › Settings › Credentials › LLM Providers and click Add provider.
- Choose the provider you want to use.
- Paste the required secret or complete the provider connection flow.
- Save the credential with a recognizable label.
Credential secret values are stored as encrypted runtime secrets. After creation, Botyard shows only safe metadata such as the label, provider, and masked key prefix.
3. Deploy your first bot
- Open User menu › Settings › Bots › Deploy Bot.
- On the Name step, give the bot a name.
- On the Template & Config step, keep Chat setup, the default: the bot introduces itself in its first chat and helps you finish its configuration. You can choose From template instead. Then click Continue.
- On the Credentials & Model step, choose where the bot's model comes from:
- Use Botyard credits: Botyard pays the model provider and the usage is billed to your Botyard credits, so you don't need an API key. Pick one of the recommended models to deploy the bot.
- Bring your own key: use an account you already have with a provider such as OpenAI, Anthropic, or OpenRouter, and the provider bills you directly. Select or add the provider credential, choose a model, then click Deploy Bot.
Botyard then shows a Setting up bot name page while it provisions the bot. This usually takes one to two minutes. When the bot is ready, click Open new chat and send your first message.
Video walkthroughs
- Deploy your AI bots (about a minute): naming the bot and setting its avatar, picking a template, choosing a provider and model, and confirming the bot is live.
- Working with your AI bots: the chat interface, thinking level and behavior settings, and assigning skills from the skill catalogue.
4. Next steps
- Host apps with Botyard when your bot creates a local web app you want to share with your team.
- MCP setup templates when your bot needs extra tools from MCP servers.
- API Authentication when you want to call the Botyard API from your own automation.
- Manage Botyard with Terraform when you want to declare bots, skills, MCP servers, and vault secrets as code.
Overview
Botyard documentation: deploy bots, connect provider credentials and GitHub, manage Runtime Vault secrets, run workforces, host apps, and use the REST API.
Provider credentials
Add LLM, web search, and integration credentials in Botyard, how they are stored, and how they are injected into outbound provider calls at request time.