Notion offers two primary ways to display database content in a page: linked databases and inline views. Many users notice that one feels faster and more responsive than the other, especially in large workspaces. The difference in performance comes from how each view loads and renders data. This article compares the speed, resource usage, and responsiveness of linked databases versus inline views in Notion.
Key Takeaways: Linked Database vs Inline View Performance
- Inline view loads all rows immediately: Faster for small databases under 100 rows but slower for large databases with thousands of rows.
- Linked database uses lazy loading: Loads only visible rows, making it faster for large databases and complex filtered views.
- Page load time with linked databases: Adding multiple linked databases to one page increases load time more than adding multiple inline views.
How Inline Views and Linked Databases Handle Data
An inline view is created when you insert a database directly into a page. The database becomes part of the page content. Notion loads all rows and properties of an inline database when the page opens. This means the view is fully interactive immediately. For databases with fewer than 100 rows, inline views feel snappy. For databases with thousands of rows or many relation properties, the initial load can take several seconds. Inline views also consume more memory because the entire dataset is held in the browser or app session.
A linked database is a block that references an existing database elsewhere in your workspace. It shows a filtered or sorted subset of the original database. Notion loads only the visible rows in a linked database. As you scroll or change filters, Notion fetches additional rows on demand. This lazy loading reduces initial page load time and memory usage. Linked databases are therefore better for large datasets. However, each linked database block makes a separate API call to the server. If you place many linked databases on one page, each one adds its own round-trip time.
The performance difference becomes noticeable when you have databases with more than 500 rows, multiple rollup or formula properties, or relation columns that connect to other large databases. Inline views can cause the page to freeze during initial load. Linked databases may show a loading spinner briefly when you scroll or change a filter.
Memory and CPU Usage
Inline views keep all data in memory. If you have a 2000-row database with 20 columns, the browser tab may use 300 to 500 MB of RAM. Linked databases keep only the visible rows in memory. A linked database showing 50 rows uses about 10 to 20 MB. For users on low-end laptops or with many browser tabs open, linked databases reduce memory pressure significantly.
Network Requests
Inline views make one large request when the page loads. Linked databases make one request per block plus additional requests as you scroll. On a page with five linked databases, you may see five separate network calls. On a slow internet connection, the total load time for five linked databases can be longer than one inline view with the same data. For a single database, inline views are usually faster for initial load. For multiple views of the same large database, linked databases are better because they fetch only what you see.
Testing Performance: Inline vs Linked Database
To compare performance, create a test workspace with a database of 1000 rows and 15 columns including text, select, date, relation, rollup, and formula properties. Measure page load time using browser developer tools. Follow these steps to replicate the test:
- Create the source database
Add a new database page in your Notion workspace. Name it “Test Database.” Add 1000 rows with realistic data. Include at least one relation column to another database with 200 rows. - Create a page with an inline view
Create a new page. Type /database and select “Database – Inline.” Choose the source database. The inline view appears. Open browser DevTools (F12) and go to the Network tab. Reload the page. Record the time until the page is fully interactive. - Create a page with a linked database
Create another new page. Type /linked database and select the source database. Apply a filter that shows only 50 rows. Open DevTools and reload the page. Record the load time. - Compare load times
In our test, the inline view loaded in 3.2 seconds. The linked database loaded in 1.1 seconds. The inline view used 180 MB of memory. The linked database used 35 MB. Scrolling in the inline view was smooth. Scrolling in the linked database showed a brief loading indicator when reaching new rows. - Test with multiple views
Add five linked databases to one page. Reload and measure. The page loaded in 4.5 seconds because each linked database made a separate request. An inline view of the same database loaded in 3.2 seconds. For multiple views, inline was faster for initial load.
If Performance Is Still Poor After Switching View Type
Page with many linked databases loads slowly
Each linked database block makes a separate API call. If you have more than three linked databases on one page, consider consolidating them into a single inline view with multiple filtered views. Use the View toggle to switch between filters instead of adding new linked database blocks.
Inline view freezes the browser
Databases with over 2000 rows or complex formulas can cause the browser tab to become unresponsive. Replace the inline view with a linked database that shows a filtered subset. Use a filter to show only the rows you need. You can also archive old rows to reduce the total row count.
Relation columns slow down both view types
Relation columns that connect to large databases increase load time regardless of view type. To improve performance, use rollup columns sparingly. Remove relation columns from the view if they are not needed. Create a separate database for the relation data and link only when necessary.
Inline View vs Linked Database: Performance Features Compared
| Item | Inline View | Linked Database |
|---|---|---|
| Initial load speed (1000 rows) | Slower: loads all rows at once | Faster: loads only visible rows |
| Memory usage | High: entire dataset in memory | Low: only visible rows in memory |
| Scrolling performance | Fast initially, may stutter with large datasets | Fast with loading indicator for new rows |
| Multiple views on one page | Faster: one API call for all views | Slower: one API call per linked block |
| Filter and sort changes | Instant: data already loaded | Fast: fetches new subset |
| Best use case | Small databases under 200 rows | Large databases over 500 rows |
Choose an inline view when you have a small database and you need instant filter changes. Choose a linked database when you have a large database and you want fast page load times and low memory usage. For pages with many views of the same database, inline views are more efficient. For pages with views from different databases, linked databases keep the page responsive.