Your AI chatbot is only as good as the knowledge you give it. This guide covers everything you need to know about uploading documents, crawling websites, and organizing your knowledge base for optimal results.
1. Supported File Formats
Chat.co supports a wide variety of file formats to make it easy to import your existing content.
Documents
Spreadsheets
Media (Pro & Business Plans)
Note: The Starter (free) plan only supports PDF, TXT, and DOC files. Upgrade to Pro or Business for images, video, and spreadsheet support.
2. Size & Storage Limits
Upload limits vary by plan. You can purchase add-ons to increase your limits if needed.
File Upload Limits
| Plan | Per File | Total Storage |
|---|---|---|
| Starter (Free) | 10 MB | 50 MB |
| Pro | 10 MB | 100 MB |
| Business | 10 MB | 500 MB |
| Enterprise | 10 MB | 5 GB |
Knowledge Base Storage
Your knowledge base storage is measured in characters (roughly 1 character = 1 byte).
| Plan | Character Limit | Approx. Pages |
|---|---|---|
| Starter (Free) | 100,000 | ~50 pages |
| Pro | 2,000,000 | ~1,000 pages |
| Business | 4,000,000 | ~2,000 pages |
| Enterprise | Unlimited | Unlimited |
Need More Storage?
Add-ons are available to increase your limits:
- File Upload Limit — +100 MB ($59/month)
- Knowledge Base Expansion — +1,000,000 characters ($10/month)
3. Uploading Documents
Adding documents to your chatbot's knowledge base is straightforward.
Step-by-Step Upload
- Go to your chatbot's Sources page
From your dashboard, select a chatbot and click
Sourcesin the sidebar. - Click "Add Source"
Select
Upload Filesfrom the dropdown menu. - Select your files
Drag and drop files or click to browse. You can upload up to 5 files at once.
- Wait for processing
Small files process in seconds. Larger PDFs may take a few minutes. You'll see a green checkmark when complete.
Tip: For best results, use text-based PDFs rather than scanned documents. Scanned PDFs require OCR processing and may have lower accuracy.
4. Website Crawling
Instead of manually uploading documents, you can automatically import content from websites.
Crawl Types
Full Website
Crawls your entire website recursively, following links to discover all pages. Ideal for documentation sites, blogs, and company websites.
Single Page
Crawls only the specific URL you provide. Perfect for individual articles, FAQs, or landing pages.
YouTube Video
Extracts the transcript from a YouTube video. Great for training videos, tutorials, and presentations.
URL Crawl Limits
| Plan | URL Crawls |
|---|---|
| Starter (Free) | 100 |
| Pro | 10,000 |
| Business | 25,000 |
| Enterprise | 25,000+ |
Need more crawls? The URL Crawl Boost add-on provides +10,000 crawls for $5/month.
Automatic Re-crawling
Keep your knowledge base up-to-date by scheduling automatic re-crawls:
- Daily — For frequently updated content
- Weekly — For moderately changing content
- Monthly — For stable content that changes occasionally
5. Organizing Your Knowledge
Sources are displayed as a flat list within each chatbot. For complex knowledge bases, we recommend organizing by creating separate chatbots for different topics.
Recommended Approach
Single Bot for Simple Use Cases
If all your content is related (e.g., product documentation), a single chatbot with all sources works well.
Multiple Bots for Complex Organizations
For diverse content, create separate bots: one for HR policies, one for product support, one for sales materials, etc.
Example Organization: HR Support Bot ├── Employee Handbook.pdf ├── Benefits Guide.pdf ├── PTO Policy.pdf └── https://company.com/hr-faq Product Support Bot ├── User Manual.pdf ├── API Documentation.pdf ├── Troubleshooting Guide.pdf └── https://docs.company.com Sales Bot ├── Product Catalog.pdf ├── Pricing Sheet.xlsx ├── Case Studies.pdf └── https://company.com/solutions
6. Best Practices
Follow these guidelines to get the best results from your knowledge base.
Use clear, descriptive file names
"2024-Employee-Benefits-Guide.pdf" is better than "doc1.pdf". Good names help with citation clarity.
Structure documents with headings
Use H1, H2, H3 headers in your documents. This helps the AI understand content hierarchy and improves retrieval accuracy.
Remove duplicate content
Multiple documents with the same information can confuse citation attribution. Keep your knowledge base clean and deduplicated.
Keep content up-to-date
Outdated information leads to wrong answers. Use scheduled re-crawls for websites and periodically refresh uploaded documents.
Test after adding content
After uploading new sources, test your chatbot with questions related to the new content to verify it's being retrieved correctly.
Pro Tip
Start with your most important documents first. It's better to have a small, high-quality knowledge base than a large one full of irrelevant content. You can always add more later.