Documentation Index
Fetch the complete documentation index at: https://docs.tryblazar.com/llms.txt
Use this file to discover all available pages before exploring further.
What is a Blazar table?
A table is the core unit of research in Blazar. Each table represents a single research question and contains the structured data that Blazar’s AI agents find on the web. A table consists of:- A name — a descriptive title you give to your research (e.g., “AI SaaS Companies in Europe”)
- Rows — one for each entity that matches your query (e.g., individual companies)
- Columns — one for each data point you asked for (e.g., website, headquarters, employee count)
- Cells — each cell contains a researched value with one or more citations

Creating a table
You can create a new table in two ways:- Click “Create new table” in the empty workspace
- Use the keyboard shortcut:
Cmd+;(Mac) orCtrl+;(Windows/Linux)
Naming your table
When you create a table, a title field appears at the top of the workspace. Click it to type your table name. A good table name describes the research question — it helps you identify the table later when you have multiple tables. Examples of good table names:- “European AI startups founded in 2025”
- “Best laptop models under $1000”
Table structure
Blazar automatically determines the table structure based on your query:| Component | What it represents | Example |
|---|---|---|
| Rows | Entities that match your query | Notion, Figma, Datadog |
| Columns | Data points you asked for | Website, headquarters, employee count |
| Cells | Researched values with sources | ”San Francisco, CA” with citation |
Cmd + 0. See Citations for more details.
You can also add rows (Cmd + [) and columns (Cmd + ]) to an existing table using keyboard shortcuts.