Suggest a Feature

I’d to request support for using variables within prototypes for animation properties such as duration and easing. We’re working to systematize motion in our design system, and the ability to define duration and easing as variables would make motion much more consistent and reusable across files and teams. This is important because these values are foundational to how motion feels in the product, and they should be treated with the same level of structure and control as other design tokens. Right now, there’s a disconnect between how motion is defined in code and how it’s represented in Figma, since animation values can’t be managed in the same systematic way. Supporting variables for prototype animation properties would help bridge that gap, improve alignment between design and engineering, and make motion easier to maintain at scale.

Hey Figma team and community,Is there any chance we could get native support for WebP and WebM exports anytime soon?Right now, exporting web assets feels a bit outdated. We’re still relying on giant PNGs or super heavy GIFs, only to manually run them through external conversion sites before sending them off to dev.Having native WebP for crisp, lightweight images and WebM for smooth UI video clips (especially with transparent backgrounds) would be a massive quality-of-life upgrade. It would make the design-to-code handoff so much cleaner.Upvote if you’re also tired of using 5 different file converters every day!

Hello, It would be more than interesting to be able to create variable at a component scope. And of course instance to inherit those variables. Why ? it would be a good way to organize variables. One file can have a lot of different components and subcomponents = a lot of variables. it would give us the possibility to ease prototype creation by preparing upfront variables in components library. it is a good way to get rid of variants. I mean variants are nice but variables are far better :). it would give the possibility to create different set of variables corresponding to different state of the components faster.

Hey there! I was also wondering if it is somehow possible to include also the layer of variant controls to be exposed or hid! I often have cases where I need to nest base components e.g. an icon wrapper providing me with the ability to select certain sizes, or color overrides and boolean props of an icon. I am now able to expose all of those or none, but I can’t restrict the users to only be able to select just some of those. What do you think about it?

Request: Through the Figma MCP, access contents of comments made in Figma Design file.Reason: Can provide context for designs, review bugs, etc.Current response from OpenAI Codex with direct link to comment:No, not the comment itself.What I can see from that link is the node it points to:file: XXXXXXnode: XXXXXXscreen: /sign-up-2 (iOS)That screen contains:nav title Sign UpEmail and Password inputsSign Up primary CTAContinue with AppleContinue with Googledivider OrAlready have an account? Sign inbottom legal linksBut I still cannot read the actual Figma comment thread or the #XXXXXX anchor content from this session.If you want, paste the comment text here and I can map it to this exact screen and tell you what change it implies. 

Every time I exit Figma and come back, the previously selected sort order is lost, and it defaults to "Last modified" – "Newest first". Previously, the selected sorting option was always preserved.

It would be amazing if there was an advanced clipping option that would allow designers to choose what edges clipped content and what edges let content run over it. I am working on complex enterprise software where I have dropdown menus with menu items with hover states as well as icon buttons in the menu items with their own hover states. And on top of that I need to show tooltips on some of the icon buttons and allow clicking to open a new modal. (There is a lot going on) We have workarounds for the overlay tooltip but the issue we’re running into is everything is in a scrollable container and that requires clipping to look decent. It would be nice to decide what edges get clipped. In our scenario we need clipping at the top and bottom of the scrollable container but the left and right don’t need it so that absolute positioned items can spill outside of the scrollable container.

It would be nice if I could quickly open pages in my design file in a new tab. Often pages contain designs that would be useful to view side by side with the current page you are working on. It would be nice if I could quickly open them in a new tab / new window instead of having to go back to files, open the file, wait for it to load and then navigate to the page.

Hi! My organization has limited the use of AI to just the protected environment Copilot 365 has to offer. I can use Copilot Studio, which provides Agents with ‘tools’ to access a range of MCP servers. Figma, sadly, is not one of them.If I attempt to add a custom MCP server for Figma to Copilot Studio, it fails on authentication. I’ve seen other posts on the forum about the issue as well. According to Copilot this issue is caused by Figma not allowing Copilot to authenticate. Since Microsoft has such a large audience using their apps, it would be great if Figma allowed Copilot 365 to also access to authenticate.Would it be possible for Figma to add/allow Copilot 365?Thanks!

Feature Request: Git-like Version Control for Figma BranchesCurrently, Figma allows us to create a branch from each file, but this limitation makes it difficult to manage complex design workflows — especially for teams working on large design systems.Git-like branching would make version control in Figma far more powerful and flexible. With this approach, designers could create multiple branches from existing ones — functioning much like environments (e.g., test, staging, production) in development.Attempt at elucidating through a flow diagramAs a design system maintainer, this would be a huge benefit. For example: Components that need additional review or testing could be safely designed in separate subbranches. Compatibility between different branches could be tested before merging changes into the main file. It would reduce merge conflicts and ensure higher quality and consistency across the system before releasing updates. Allow for other users to expand on branches through t

The gapVariable mode inheritance only works within a single collection. A mode cannot be derived from anything: not from a variable, not from an alias, not from a component property. This means that in any composed component, the mode of a nested component must be set by hand on every instance. Concrete caseAn Alert has two sizes, driven by modes in an `Alert size` collection. It contains a Button and an Icon, each with their own size collection. What I need to express is a mapping:| Alert size | Button size | Icon size || --- | --- | --- || small | Small | 14 || large | Default | 18 |There is currently no way to state this relationship anywhere in the file. Setting the Alert to `large` leaves the nested Button at whatever mode it happens to carry. Why the existing workarounds don't hold upOne shared collection.This requires the scales to align on two axes. They don't. My icons have five steps (12/14/16/20/24) and my buttons have three (Default/Small/Large). Forcing both into one mode

In Dev Mode, when selecting an image, you can see the original filename of the placed source image and download it by 1) clicking on the name in the layer properties, or 2) in Assets.However, when downloading, a strange filename is suggested that doesn't match what we see in the layer properties.Question 1: Is it possible to fix this so that the source image is downloaded with the original name?Question 2: Is it possible to bring this functionality into the main editing mode? It's sometimes very useful when working with designs, and it's sorely missed.

Would love it if you could paste a screenshot or image into a sticky note.It’s obnoxious to paste an image onto the whiteboard, move it to align with a sticky note, and then group the two elements together. Takes more time than it should :(

The default settings (at least in my Figma) load a percentage for letter-spacing, but this is not allowed in CSS. For a less avid programmer this may be confusing, when trying to get it to work. Either set the default to a px value, or add a warning inside the Inspect code

new AI agent “describe your idea” icon/text box feature when i click on an instance/frame in my design is a bit intrusive - little star icon gets in the way when i’m trying to adjust my design manually. maybe limiting it to the right click menu or simply creating a new hotkey would be better? also, telling this agent what to do is a miserable experience that usually just frustrates me back into doing the design myself as it is literally faster than typing it out. have seen a lot of feedback similar to this as well here.

Figma is no longer used only by designers and developers.In many real-world projects, designers share Figma files or prototypes with external stakeholders such as clients, managers, marketing teams, legal teams, and other non-design users for review and approval.These viewers did not necessarily choose Figma themselves, and many of them may not be familiar with Figma or comfortable using an English interface.This creates unnecessary friction in the review process.The intended journey should ideally be:Open shared link → Review the designBut in practice, it can become:Open shared link → Understand an unfamiliar English interface → Figure out how to navigate the prototype viewer → Review the designThere is another common scenario where this becomes even more problematic.After a client receives a prototype or shared file, they may need to share it with other members of their internal team. At this point, they have to interact with Figma's sharing interface themselves.This interface involv

When overriding a single element inside a slot in a component instance, Figma marks the entire slot as overridden. As a result, updates made to the main component no longer propagate to any elements inside that slot — even if only one element in the slot was actually changed. Expected behaviorOnly the modified element should be treated as overridden.All other elements inside the slot should remain linked to the main component and continue to receive updates.

Hello,some feedback from my side regarding Folders, I like the options of a nestable folder, but I find it hard to navigate to the files I’m looking for. The folders are named, but the file thumbnails that were visible with Projects were a huge readability feature that I’m really missing now.I was searching for it in settings, but couldn’t find a way to set it. I actually added a description to a folder, but did not actually see not one place that the description was shown? This part is especially confusing.If I could make an improvement suggestion to a current implementation it would be to have a setting to show or not to show some folder contents thumbnails, similar to the previous implementation.Fingers crossed this can be done!

Hello,When a user holds both admin and regular user roles on Figma, notifications get mixed together: license requests, AI budget overages (admin) and comments (user) all appear in the same feed, which makes them hard to read through.We currently have 3 tabs: All / Request / Unread.Would it be possible to add a dedicated tab or filter for comments, so that admin-related items (license requests, AI credits) can be clearly separated from day-to-day usage items (comments)? This would make it easier to prioritize and ensure nothing is missed depending on the context we're reading in.Thank you in advance for your feedback on the feasibility of this improvement.

Allow designers to create an AI-powered project space where the agent can use project-specific sources and default memory to understand the design context.Designers can upload or connect sources such as brand guidelines, design systems, UX documentation, product requirements, or tone-of-voice guidelines. The AI agent uses these sources as the primary context when generating, reviewing, or modifying designs.Inside the AI agent panel, designers can configure the project's context and choose which sources the agent is allowed to use.Example ScenarioA designer creates a Nike Brand Campaign project in Figma.They add:Nike brand guidelines Typography guidelines Color specifications Logo usage rules Brand voice guidelinesThe designer then sets the AI Agent's project context to:Sources✓ Nike Brand Guidelines ✓ Project Design System ✓ Default Figma Memory ✕ Other ProjectsThe agent can now generate or critique designs based specifically on the selected project sources.For example:"Create a campai

Describe the problem your experiencing and how your idea helps solve this When you create an autolayout frame, it always defaults to 10px. Many design systems call for an 8px grid and in many cases, users don’t want ANY padding especially where edge-to-edge designs are concerned. Ask questions to bring the community into the conversation What are everyone’s thoughts about the following solutions?: Default autolayout padding based on the big nudge amount in your settings Allow for tokenization of padding/margin grids and offer up autolayout parameters based upon those tokens Allow users to override the padding to 0px at the creation of the autolayout frame by using an alternative key command like ctrl+shift+a (command on mac)

Dear Figma Team,I would like to propose a feature that could be useful for UX designers and researchers. Problems I identified during my research:CJM, JTBD, empathy map templates, and other research methods are scattered across different tools — Figma, Google Docs, Notion, Miro. It takes 15–20 minutes to find the right method, and there is no centralized storage.Difficulty choosing the right research method.Fragmented workflows. After analyzing these problems, I created an application called UX Logic, which addresses them:Adding a library of methods and research materials, with the ability to add custom libraries and research materials. AI recommendations on the home screen. The AI suggests: “Structure insights,” “Validate hypotheses,” and “Build a CJM.” A unified workflow: Home - Projects - AI Chat - Library - Profile. All tools are in one place, without switching between 5+ applications. How can you use this application?Here is an example of one scenario:You paste interview notes int

When creating nested components, I want to lock the contents of a child components slots. For example, I make a generic “menu” component with slots. So then I make a specific “File Menu” using that generic menu component. That’s great. But when I turn that File Menu into a component, the contents are still editable in instances of the File Menu. I don’t want that. I don’t want consumers of the File Menu turning it into an Edit Menu. Standard Lock isn’t what I want, because I do want the user to be able to apply regular overrides, just not use the more drastic “Slot” features.So my request is for a feature that allows a main component to disable child slots.

Hello,I always enjoy using Figma Design and FigJam.I use FigJam for organizing Minecraft addon development and system design, and as projects become larger, I’ve started encountering difficulties balancing “the amount of information inside nodes” with overall readability.Currently, FigJam is extremely effective for:mind mapping organizing ideas visualizing dependencies structuring systems and architecturesHowever, as projects scale up, each node tends to require a large amount of detailed information, such as:identifiers recipes geometry references technical notes implementation status tags JSON-related information various propertiesIf all of this information is displayed at all times, the overall board quickly becomes difficult to read and manage.Because of this, I would love for you to consider adding:“Expandable / collapsible node functionality”For example:Normally, a node would display only a simple summary Pressing a button would expand detailed information A larger text area or s