I have regular problems with rich text field not being able to store long texts (like long ai chats or analysis)
Please let me know if this information is correct:
Fibery’s rich text fields are limited to ~2.7 MB (suitable for most content but not very long texts),
and attachments support up to 500 MB but lack native text extraction in automations or AI workflows.
Issue:
This creates a notable gap for document-heavy or automation-reliant use cases, such as knowledge management, report processing, or AI-driven insights, making Fibery less competitive against tools like Notion or Airtable.
Severity depends on needs:
serious for scalable, integrated workflows (requiring workarounds like content splitting, external APIs, or Zapier integrations);
minor for lighter tasks.
What is the plan of the Fibery team to address this?
I have used AI to try to understand just how much content “2.7MB” would be, since it’s a very odd way to think about text content to me (not sure how you arrived at that number, kind of curious). I know you’re a fan of AI for writing so:
For Markdown Format (2.7 MB):
Markdown’s overhead for basic formatting is minimal. Therefore, it’s very close to pure text.
Characters: Approximately 2,831,155 characters
Words: Approximately 566,231 words
Pages: Approximately 1,132 pages
For HTML Format (2.7 MB):
HTML, even with basic formatting (paragraphs, bold, italics, headings), has significant overhead from tags and structure. We’ll estimate an overhead of about 25% of the file size dedicated to markup, leaving 75% for content.
Effective Text Content: ~2.025 MB
Characters: Approximately 2,123,066 characters
Words: Approximately 424,613 words
Pages: Approximately 849 pages
Now resuming my own writing: even in HTML this seems like an insane amount of text to want to put into a Rich Text. Are these assumptions about content length (above) incorrect? If not what on Earth are you trying to put in there? I am surprised and confused but look forward to understanding.
We collect feedback and use cases, then prioritize based on overall impact. So far the 2.7MB limit per document hasn’t come up often as a major pain point.
If it were a serious issue, we’d expect to see more feedback about it.
Do you have any limits regarding the Document’s size?
There is a limit on the size of the document - 2.7 Mb. But this size can’t be calculated simply by the number of characters - it includes inner document metadata.
So there is no limit to the number of characters on the same page before breaking down, but once the document is full - it will reject any new typings with a warning.
“But this size can’t be calculated simply by the number of characters - it includes inner document metadata.”
I want to copy paste an AI chat (not the build in Fiber AI) in the rich text field, in which I develop a script. So this often contains a number of version of the script in the same chat. I use premium AI with high token context windows.
To have an entire AI chat is very important for analysis.
Sometimes I have long AI chats writing and adapting acadamic papers, which can be also very long chats (50 to 100 turns)
–
I like to prevent using third party tools to pull in such data in fibery. So either I put it in a rich text field, or in an attachment if fibery can fetch the text of that in the future, using automations.
Thank you, that makes some sense. But still seems like 2.7MB of text should be enough for most uses, and where it’s not, AI analysis or summarization or something can perhaps be done separately in whatever automation is getting it into Fibery in the first place.
Why I encounter the limitations of a rich text field (and need for attachments not to be accessible by automatins) is likely because I’m developing workflows that are not just a simple team workflow.
I think that this issue grew out of the the past idea of Fibery being a tool while now emerging into an operating system for a company, which means that commonly used large documents or resources need somewhow be processable by Fibery:
So there is no integrated solution in Fibery for, e.g.:
Meeting Minutes
HR Policies
Meeting Transcripts
Project Specifications
RFPs
Employee Handbooks
Market Analysis
Research Reports
Technical Documentation
User Manuals
Legal Contracts
Annual Reports
Financial Statements
Data Exports
CSVs
Spreadsheets
Proprietary Algorithms
Source Code Repositories
Where to store these, when you do not want to rely on external tools?
These you want NOT to split up in multiple fibery entities, and can be too big for a rich text field,
And as fibery entity attachment they are not accessible/workable with (unless through external API).
So confusing. There are very many solutions for almost all of those things. Almost none of them need a monolithic 800+ page Rich Text field. Fibery’s own User Guides prove that Fibery can certainly be used for that, and shows a way to do it!
Fibery is explicitly not trying to solve this. Don’t expect it to.
It kicks in much faster than what you repeat as 800 pages. I’m fine with any other solution, i’m not married with rich text fields. But attachments dont work either.
But apart from that, to me its not confusing at all - The entire list are very common forms of content that are large and commonly part of orgaizations.
I’m not pushing them to be part of rich text fields.
The title of this topic indicates that I find it difficult to match long content with Fibery in its entirety - there appears no support for that, other than:
‘find out yourself how you handle that’ - we don’t support it.
Suggestion
So it would be good for Fibery to assist users faced with this ‘intentional limitation of fibery’ to explain how such integration should happen, from the perspective of Fibery as a product.
Am I now formulating it in a better way?
If the answer of Fibery will be a list of external API apps, thats fine with me; I accept that, and I will move on.
If they don’t want to do that, then I wonder why they call it ‘your company operating system’, which assumes coverage of ways to integrate common workflows, either internally or externally.
To the best of my recollection, we have not had any other customers complaining that the size limits for rich text have prevented them from using Fibery for their company use cases.
Even if the Fibery size limits work out as less than 800 pages per doc in practice, say 500, or even 250 pages, we just don’t hear from users who experience this as a problem, especially since docs/entities support nesting etc.