What chunk size should I use?

For most RAG applications, 200-500 tokens per chunk works well. Smaller chunks improve precision but may lose context; larger chunks preserve context but reduce retrieval accuracy.

Why does overlap matter?

Overlap ensures that information at chunk boundaries isn't lost. A 10-20% overlap helps maintain context continuity between adjacent chunks.

What chunk size should I use?

For most RAG applications, 200-500 tokens per chunk works well. Smaller chunks improve precision but may lose context; larger chunks preserve context but reduce retrieval accuracy.

Why does overlap matter?

Overlap ensures that information at chunk boundaries isn't lost. A 10-20% overlap helps maintain context continuity between adjacent chunks.

Gratuit

Aperçu des blocs RAG

Collez le contenu d'un document et visualisez comment il est divisé en blocs pour la récupération RAG. Ajustez la taille des blocs, le chevauchement et les jetons minimums.

100 % gratuitAucune inscription requiseFonctionne dans votre navigateur

# Getting Started with LaunchChat

LaunchChat is an AI-powered support widget that transforms your documentation into an intelligent chatbot. It answers questions based strictly on your content with citations.

## Installation

To get started, sign up for a free account at launchchat.dev. Then connect your Notion workspace or upload your documentation files.

### Connecting Notion

1. Go to the Setup page in your dashboard
2. Click "Connect Notion"
3. Authorize LaunchChat to access your workspace
4. Select the pages you want to sync

### Uploading Files

You can also upload Markdown, PDF, or DOCX files directly. Each file will be parsed and chunked automatically for optimal retrieval.

## Configuration

After connecting your content sources, configure your widget settings:

- **Confidence Threshold**: Set the minimum confidence score for answers (default: 0.7)
- **Max Output Tokens**: Control response length (default: 500)
- **Refusal Message**: Customize what the bot says when it can't answer
- **Show Related Articles**: Enable article suggestions below answers

## Embedding the Widget

Add the widget to your website with a single script tag:

```html
<script>
  window.NotionSupportConfig = { widgetId: 'your-widget-id' };
</script>
<script src="https://launchchat.dev/widget.js"></script>
```

The widget will appear as a floating chat button in the bottom-right corner of your page.

Jetons maximums : 400

Chevauchement : 50

Jetons minimums : 30

1,404 / 100,000 chars

À propos de RAG Chunk Previewer

See exactly how your documents will be split into chunks for retrieval-augmented generation (RAG) systems. Adjust chunk size, overlap, and minimum token thresholds. Visualize chunk boundaries, token counts, and overlap regions to optimize your RAG pipeline.

Questions fréquemment posées

Plus d'outils AI Tools

SaaS Profit Calculator

Calculate your SaaS monthly profit and unit economics.

Input your SaaS metrics — monthly active users, paid conversion rate, subscription price, and per-user costs — to calculate MRR, ARR, gross margin, monthly profit, ARPU, LTV, and break-even point. A practical tool for indie hackers and SaaS founders.

Text Diff

Compare two texts and highlight the differences.

Website SEO Checker

Run a 30-point SEO audit on any webpage.

Voir tous les outils

C'est exactement comment LaunchChat traite vos documents. Connectez Notion et déployez en 5 minutes.

Aucune carte de crédit requise

Visualizing chunks? LaunchChat handles the entire RAG pipeline — from chunking to embedding to retrieval to answer generation.

Free tier available. No credit card required.

Try Free