Help Us Improve Tiptap: Share Your Feedback

[bdbch]

Hey Tiptap community,

We want to make Tiptap better over the next year, and the most valuable input for that comes from the people actually using it in real projects.

I’d love to hear your honest feedback on things that are currently not working well or are causing friction for you. In particular, I’m interested in:

• Pain points & blockers
  Bugs, edge cases, or limitations that slow you down or make certain use cases hard or impossible.

• Documentation gaps
  Parts of the docs that are unclear, misleading, outdated, or simply missing.

• API & code quality feedback
  Areas where the API feels awkward, inconsistent, or harder to use than it should be.

• Project & DevRel experience
  How we handle issues, PRs, discussions, releases, and communication in general. What works well, and what doesn’t?

A quick note on scope: this discussion is intentionally focused on the open-source core of Tiptap. Pricing or paid extensions aren’t the main topic here. We want to understand what you need from the OSS side to confidently build and ship your apps. Our goal is to empower developers to extend and customize Tiptap themselves when needed, while keeping the core solid, well-documented, and pleasant to work with.

This is not a promise to immediately implement everything or to prioritize individual feature requests. The goal is to identify patterns, recurring problems, and areas where we can meaningfully improve the project for everyone.

If you have feedback, please reply to this thread. Concrete examples, links to issues, or short explanations of your use case are especially helpful.

Thanks for helping us make Tiptap better.

And since we’re close to the end of the year: we wish you a peaceful Christmas time with your family and friends, and a great start into the new year.

[slmnsh]

The documentation around Node, Mark and other primitive sucks. there is no clear explanation of different use cases.

You need to go through existing tiptap extension to figure out how to do stuff.

[bdbch]

Do you have specific cases - or something that is under-explained? (Or more, what exactly would you expect to learn/know about the different types of primitives).

[slmnsh]

• There is no clear explanation for all the utilities that are available in list or table extensions we need to go through Github to find out about them.
• Docs should contain commonly used snippets like updating attrs of selection, getting selected table cells and operating on them etc.
• In ProseMirror, you can easily setup dispatchTransaction on editor view but in tiptap you have to jump through various hoops:

onCreate() {
  // @ts-expect-error tiptap doesn't expose dispatchTransaction
  const dispatch = this.editor.dispatchTransaction.bind(this.editor)
  this.editor.view.setProps({
    dispatchTransaction: withSuggestChanges(dispatch),
  })
}

[bdbch]

In ProseMirror, you can easily setup dispatchTransaction on editor view but in tiptap you have to jump through various hoops:

This is now supported on Extension level via the new dispatchTransaction option.

[Chaddeus]

Loving streamContent, but it's weird that it's trapped in the Start plan. Take streamContent out of the Start plan. It's not AI, it's functionality to stream text into the editor chunk by chunk.

Another suggestion: add a built-in way to mark/label streamed text (via streamContent) so we can tell it apart from what users type themselves.

Thanks for the great work; keep crushing it! 🚀

[bdbch]

That's something I already listed and want to do. The actual functionality to just stream content while still working with incoming transactions is something I'd love to support by default.

[Chaddeus]

Fantastic. I worked with Nick Perez back when streamContent was in beta to help him find bugs and smooth out use cases.

I know you didn't want to discuss costs... but if I may. An important reason I jumped in here to add feedback is because I only use this function and none of the other things in Starter. I've been on the old developer plan for years and am now getting forced into Starter to keep using streamContent.

Mentioning it to give you some food for thought when considering to move this function out of the Start plan.

Aside from that, it would be awesome if there were some examples, UI components, etc. using Svelte in the docs (they seem to be React and Vue heavy). I've never really enjoyed using React, especially once Svelte 3 arrived. Svelte 5 is fantastic.

Thank you. 🙏

[vinodkumars]

Please make AI toolkit cheaper and acessible without going through the Sales team. Especially for AI toolkit features that do not depend on Tiptap cloud like Review changes should be part of the Start plan or OSS.

[tapsuns5]

I would love for list nodes to still have capability to be dragged and dropped as individual nodes. They're nested right now so grouped together when trying to drag around.

I am going to look at re-doing this myself with a custom extension and can send over if interested.

[bdbch]

Good one. For 4.0.0 we would love to make the DragHandle extension more flexible when it comes to nested nodes. The biggest problem I ran into was making it flexible enough for a wide userbase with some kind of ruleset that allows you to deactivate it for specific subnodes.

[bdbch]

Actually it was faster than expected. Nested Drag Handles are out since 3.17.0 :)

[tapsuns5]

Oh wow, nice! Beat me to it lol.

Need to upgrade to v3 to use? Still on v2..

[bdbch]

Yes, it's only on Tiptap v3.

[Halftone-Digital]

Pages is really impressive, but there isn’t much documentation about what’s possible.

We’re building a product whose main use case is turning written entries into printed books. We’re not relying on Pages to format content for print, but we do like the idea of letting users write inside a paginated editor so it feels closer to the final result.

Right now, two things stop this from feeling seamless:

1. Header and footer customization
   Each entry has a title that we display on the first page of the entry. Ideally, we could show that title in the editor header without storing it inside the entry content itself. We can’t do this until we can have a different header for the first page, and another header for all following pages.

2. Inline floating images with text wrap
   We let users place images inline so an image can sit on the left and text wraps around it. This doesn’t seem possible with the Pages extension today. For context, we’re using a custom image node so users can adjust appearance and toggle a caption.

[bdbch]

1. Header and footer customization is currently being worked on and is coming! I can't give you an exact ETA on that.
2. Can you describe a bit more what happens on your end and what the expected behavior is? It would be good to create issues for this case and link it here. Issue reports in discussions will go under.

Maybe @alexvcasillas can help out with more details here.

[alexvcasillas]

Sorry for the late response, I was on parental leave, then vacation, and then sick. It's been a ride the end of 2025 and the beginning of 2026 😅

Headers and footers are now revamped in the latest alpha version and they have their own editors where you can work with them as a usual tiptap editor. You'd still need to store the header and/or footer separately from your main editor's content because there's no other way to do it as they're 3 independent editors. You can now also choose "Different First Page" and/or "Even/Odd pages" for your headers and footers. I'm afraid I don't quite undertand your worry about: "we could show that title in the editor header without storing it inside the entry content itself" if you cold elaborate that would be very helpful.

Related to inline floating images with text wrap: your extensions should work fine when using the pages extension, if you could provide a reproducible demo or snippet that would also be helpful to understand your issue here :)

[SawkaDev]

I know you specifically said this is for OSS topics, but I can not stop my self from mentioning a few of these since I do not know how else to give some of this feedback.

1. AI Extensions and Toolkit Pricing: I was sad to see that AI Changes, AI Suggestion, and several other plugins were deprecated in favor of AI Toolkit. Architecturally, I’m impressed with what the Toolkit offers and your vision for it. However, the pricing model doesn’t make sense from a developer’s perspective. If Tiptap aims to be the go-to AI-native editor for developers, I’d expect more developer-friendly pricing for the Toolkit. It seems AI Toolkit is intentionally being limited to users who can pay $$$ each month for it. I avoid using deprecated packages in production (AI Changes, AI Suggestion), and since the Team plan doesn’t currently add much value for me (I dont need any cloud services or anything, I only was on Team for extensions) without these, I plan to downgrade to Start soon. I say this for feedback, I’ve already been able to create my own replacements for the deprecated extensions that work for my use case, and I have a feeling that others who were on Team plan specifically for support of the now deprecated packages may do so as well. Which I do not like. I would rather stay at my current Team plan and use supported Tiptap packages and the AI Toolkit at my hands. I’d love to fully adopt AI Toolkit, but at its current cost, it’s not financially feasible.

2. Pages Extension and Document Export: I understand it’s complicated, but for the Pages extension, especially with DOCX export (and the upcoming PDF export), I’d love to see true WYSIWYG fidelity. If content in the Tiptap editor appears on a specific page, it should remain on that page after export. Maybe the demo in the docs isn’t applying CSS correctly, but ideally, if a paragraph starts on page 1 in Tiptap and another on page 2, they should export to those same page positions in both DOCX and PDF.

3. Performance and Profiling: I saw a LinkedIn post mentioning upcoming profiling or performance tools for Tiptap. I manage a large editor setup with many custom and heavy nodes, so I’d love to see a more detailed, production-ready performance profiling tool. That would be an incredibly helpful addition. Also, similar to Tanstack Query for instance with the ReactQueryDevtools, it would be helpful to have an out of the box "TiptapDevTools" component that exposes a lot of important details relating to the current editor, performance etc.

4. List Style Options: Native support for various BulletList and OrderedList styles like disc, circle, square, lower-alpha, and upper-roman would be great to have. It wasn’t hard to customize, but built-in support would be convenient and consistent with modern text editors.

5. Font Family Persistence: I remember reading about font handling but haven’t found a good solution yet. In editors like Google Docs or Word, when a user sets a font, then adds text and empty paragraphs, the chosen font remains active when returning to those empty paragraphs. In Tiptap, the font currently resets to the default, which breaks the writing flow a bit.

6. Tiptap Components Modularity: Although I really like the design and thought that goes into your components, they feel too heavy to import selectively. For example, if I only need the Table UI component, installing it still pulls in 100+ files. It would be great if these components were more modular so developers could integrate individual parts without needing the entire infrastructure. Maybe I’m missing something, but a lighter, composable option would be amazing.

Overall though, I love Tiptap and what you have done! I have been a user for a few years now and will continue to be one!

[bdbch]

Thanks for that extensive list. I'm sadly the wrong one to talk about pricing but will forward it to the team. For the other points I'll try to create actionables.