Project Settings

Configure your LILA project settings, find your API key, and set up security. This page covers everything from basic project info to security settings, white-label branding, and connecting your own AI provider like OpenAI or Anthropic.

Navigation: Dashboard → Select Project → Settings

Project Settings Tab

Manage basic project information and control project status.

Project Information

Update your project’s name and description to help identify it in your dashboard.

  • Project Name - A descriptive name for your project
  • Project Description - Optional detailed description of your project’s purpose

Project Status

Control whether your project is active or paused using the toggle switch.

  • Active - Widget and API are fully operational, queries are processed normally
  • Paused - Displays red “PAUSED” badge, queries are blocked until reactivated

Note: Pausing a project does not cancel your subscription. Your current subscription continues until the cycle ends without auto-renewal. Data and configuration are preserved.

Delete Project

Permanently delete your project by clicking the “Delete Project” button. You must type the exact project name to confirm deletion.

What happens:

  • Project moved to trash (can be restored from trashed projects list)
  • Chat history retained for 90 days
  • Schema and configuration preserved
  • Widget disabled immediately

Setup & Integration Tab

Find your LILA API key and widget embed code. Get your API credentials and widget integration code here.

Security Modes

Secure your AI chatbot by choosing how end-users authenticate with your widget:

Open Mode - No authentication required

  • Widget works immediately without any backend integration
  • Suitable for public-facing applications
  • Get API key and copy widget code to embed

Standard Mode (JWT) - Token-based authentication

  • Users must provide valid JWT tokens
  • Your backend generates tokens with user context
  • Requires Client Secret Key for token signing
  • Includes code examples for JWT generation

API Credentials

API Key - Unique identifier for your project

  • Copy button for easy integration
  • Regenerate button (warning: closes all active sessions)

Client Secret Key (Standard mode only)

  • Used to sign JWT tokens on your backend
  • Keep this secure and never expose to frontend
  • Copy button for safe storage
  • Regenerate button available

Widget Code

Ready-to-use HTML snippet to embed the LILA widget on your website. Includes:

  • Loader script from embed.getlila.one/loader.js
  • Widget configuration with your API key
  • Automatic version management and caching

Code Examples (Standard Mode)

Language-specific examples for generating JWT tokens:

  • Backend integration snippets
  • Token generation with proper claims
  • Security best practices

Access Control Tab

Restrict access to your AI chatbot using IP and domain restrictions.

IP Whitelist

Restrict widget access to specific IP addresses or ranges.

  • Type IP addresses and press Enter to add
  • Use wildcards for ranges (e.g., 192.168.1.*)
  • Delete all entries to allow access from any IP
  • Separate save button for IP whitelist

Examples:

  • Single IP: 192.168.1.100
  • Range: 192.168.1.* (allows 192.168.1.0-255)

Domain Whitelist

Control which websites can embed your widget.

  • Type domain names and press Enter to add
  • Use wildcards for subdomains (e.g., *.example.com)
  • Delete all entries to allow embedding from any domain
  • Separate save button for domain whitelist

Examples:

  • Single domain: example.com
  • Subdomains: *.example.com (allows app.example.com, shop.example.com, etc.)
  • Development: localhost or 192.168.1.1:3000

White Labeling Tab

Customize the branding and appearance of your chatbot to match your brand identity.

Requires: White-Label add-on purchase ($250/month)

Quick Overview

The White Labeling tab lets you:

  • Replace LILA branding with your company name and logo
  • Customize colors for light and dark themes
  • Adjust layout - position, size, and drawer direction
  • Change all labels - buttons, placeholders, status messages
  • Control attribution - hide or customize “Powered by” footer

Configuration

  1. Enable the Master Toggle at the top
  2. Configure branding (tool name, icons, colors)
  3. Adjust layout and labels as needed
  4. Click Save Configuration

For complete configuration guide, use cases, and best practices, see the White-Label Add-On documentation.

Disable White Label: Opens confirmation popup to clear all customizations and revert to defaults.

AI Provider Tab

Use your own OpenAI or Anthropic API key with LILA. Configure your own AI provider instead of using LILA’s default service (Bring Your Own Key / BYOK).

Provider Selection

Choose from supported AI providers:

  • Default - Use LILA’s AI (no charge for AI processing)
  • OpenAI - GPT models (gpt-4, gpt-3.5-turbo, etc.)
  • Anthropic - Claude models (claude-3-5-sonnet, etc.)
  • Google - Gemini models
  • Ollama - Self-hosted models

Configuration Fields

When a provider is selected, configure:

API Key - Your provider’s API key (encrypted, never exposed)

Model - Specific model name to use

  • OpenAI: gpt-4-turbo, gpt-3.5-turbo
  • Anthropic: claude-3-5-sonnet-20241022
  • Google: gemini-1.5-pro
  • Ollama: Model name from your server

Base URL (Ollama only) - Your Ollama server endpoint (must end with /v1)

Temperature (optional) - Controls randomness (0-2)

  • 0 = Deterministic, consistent responses
  • 1 = Balanced (default)
  • 2 = Very creative, more random

Actions

Test Configuration - Verify your credentials work before saving

Save Configuration - Apply and activate your AI provider settings

Clear LLM Configuration - Remove custom provider and return to LILA’s default AI