We’re super excited to share that starting with v1.113 we’re rolling out data tables (beta) to all plans.
Since the very beginning of n8n we’ve heard many of you mention the need for a proper table inside n8n to store data between workflow executions without needing to switch platforms or setting up credentials and now it’s finally here.
With data tables you can:
Save specific data from your workflow runs
Keep data around between multiple executions
Avoid duplicate runs by tracking execution status
Store reusable prompts for different workflows
Collect evaluation data for your AI workflows
Do lookups, merges, enhancements…
…and honestly, probably 100 other creative things we haven’t even thought of yet
To make sure your instance stays performant, we’ve set a 50MB limit for everyone. If you’re self-hosting (and know what you’re doing), you can change that via the ENV variable N8N_DATA_TABLES_MAX_SIZE_BYTES
Upgrade to 1.113, give data tables a spin, and let us know what you think! What’s missing? What would make it even more useful for you? We’re really curious to hear your ideas and thoughts!
This is absolutely awesome to see, I can’t wait to use these! It’s probably been my number one frustration that saving even a small amount of data between executions for all sorts of purposes requires either integrating PostgreSQL and dealing with schemas, using a third party database or API like Supabase (as handy as they are), or using variables that are powerful but are somewhat clumsy to instantiate and track since they only work in Code nodes and only save data for production executions, making testing hard. Hoping data-tables makes a ton of these things easier! Probably won’t run the new version until it’s in final release rather than pre-release, but this is awesome to see!
This is really great - when I migrated from “the other platform” almost 4 years ago, I really felt the pain of not having a simple in-app data storage. I played around with Data tables this weekend and it’s just SUCH a good and fast experience! Kudos to our Product and Engineering teams
IMPORTANT NOTE: There is an issue with very large SQLite databases that is causing instances to slow down. Out of an abundance of caution, we are unfortunately removing version 1.113.0 until we fix this issue. We hope to have this released again with a fix within the next couple of days.
It´d be cool for self hosting to be able to add a second DB, where n8n pulls the data from. So one could have performance without having to set up each time a postgres connection.
A patch was released earlier today. It has now been tested and we have high confidence. Please update to 1.113.1 (which is still in beta) to try this feature.
Does this mean we can share data between multiple workflows now? This would make splitting up complex workflows across multiple workflows so much easier.
Very happy to see this feature. I’ve been testing it out, and was commenting feedback on Reddit but someone in the Discord server said the forums is the best place to post this instead. Here’s my list(so far)
When going to the Data Tables tab, “Create Table” is not default on the upper right button, it’s defaulted to “Create Workflow” instead.
Cannot change the data type after a column is created.
Cannot set any of the column’s as primary or unique such as the ID column (To prevent duplicates)
For some odd reason, setting a column data type to “number” then pushing data from JSON array into the table, physically opening the table and looking at the rows, the numbers in the “number” data type column are not all together. For example “29683389” shows in the table as “29 683 389”. This isn’t a one off either, ALL rows exhibit the same behavior and ALL columns set as “numbers” too.
Table page can only show 50 rows per page. Which I understand is probably for performance reasons. However, there really needs to be a “search” function for the table to search for data.
Love this list @compaholic, thanks so much for sharing it.
1, 2 and 5 are all planned. For (4), I think that is just a highlighting to make it easier to read that it’s actually 29M. So the number should still be correct.
To make sure your instance stays performant, we’ve set a 50MB limit for everyone. If you’re self-hosting (and know what you’re doing), you can change that via the ENV variable 
Upgrade to 1.113, give data tables a spin, and let us know what you think! What’s missing? What would make it even more useful for you? We’re really curious to hear your ideas and thoughts! 
Read more about the data tables in the docs here.
