Is Figma Make any good?

Mixed verdict. Figma Make is the fastest AI design tool we tested (under 60 seconds per prompt, three screens generated in two prompts). The catch: the output quality looked first-year-student in our hands-on test. Generic typography, weak hierarchy, layout decisions that did not commit to the brief. As a fast first pass that you refine manually inside Figma's editor, it earns its place. As a tool that ships client-ready output on its own, it does not.

How do you use Figma Make?

Inside Figma, open the Make panel. Type a prompt describing what you want. Figma Make generates frames, components, and (optionally) an interactive prototype based on the prompt. You can then edit the output manually in Figma's editor, ask for follow-up changes via prompt, or copy individual elements into other Figma files. Standard Figma editing tools work on the output the same way they work on manually built frames.

No. Figma Make requires AI credits inside a paid Figma plan. Figma's free tier does not include AI credits in any meaningful amount. For real evaluation, plan to be on Figma Pro or higher and to spend AI credits per generation.

What is Figma Make's pricing?

Pricing is bundled with Figma's plan tiers and AI credit allotment, not sold separately. Figma Professional at $16 per seat per month includes 3,000 AI credits per month. Heavy AI use can exceed that allotment, in which case you upgrade or buy additional credits. The pricing model is closer to OpenAI's API than to a flat subscription: you pay per generation against a monthly credit pool. Plan budgets accordingly for serious work.

What's a good Figma Make alternative?

Depends on what you need. For higher visual output quality on the same brief: dMaya (with Claude Opus 4.7 or Sonnet 4.6) produced markedly better output in our test. For a free option with rough output: Google Stitch. For coding-first prototyping: Bolt or Lovable produce running apps from prompts. The full 6-tool comparison with videos lives at /blog/ai-prototype-tool-2026.

Is Figma Make better than Google Stitch?

Different jobs. Stitch is faster, free during Labs, and produces rough output that is honest about being rough. Figma Make is slower than Stitch, costs AI credits, and produces output that markets as polished but in our test came out closer to Stitch's quality than to dMaya's or Bolt's. For exploration on a budget, Stitch wins. For staying inside Figma, Figma Make wins. For shipping output that does not need heavy refinement, neither is the right pick.

Will Figma Make replace traditional Figma design?

Not for serious design work, at least not in 2026. Figma Make is an additive layer on top of Figma's existing tools, not a replacement. Designers using Figma Make productively use it as a rough first pass, then refine manually with Figma's traditional design tools. Output as-is is rarely client-ready.

Can Figma Make integrate with my design system?

Partially. Figma Make works with components and styles you have in your Figma file, so existing design tokens influence the output. Quality of integration is uneven; in our test, the output did not consistently respect color or spacing rules from the brief in the same way a traditional Figma session with manually-applied tokens would. Treat design system integration as an aspiration of the tool, not a guarantee.

Should I use Figma Make for a client project?

Only as a first-draft tool, with mandatory manual refinement before client delivery. Output as-generated tends to look generic and uncommitted, which is the wrong signal for a client review. The realistic workflow: use a stronger vibe design tool (dMaya, Stitch, Claude Design) for the first pass that goes to the client, and use Figma Make for internal iteration where output quality matters less than speed.

Figma Make · 2026 Hands-On Review

Figma Make Review (2026): Hands-On Test, Honest Take

Q: What is Figma Make?

Figma Make is Figma's AI design tool, launched in 2025, that turns text prompts into Figma frames and interactive prototypes. It lives inside the broader Figma surface, uses Claude under the hood (the model is not user-selectable), and integrates with Figma's existing design system, components, and AI credits. It is Figma's response to the vibe design category Google Stitch popularized in March 2026.

Dhairya Purohit

Builds dMaya. Ships AI design workflows in real client work.

Published May 7, 2026

Figma Make was Figma's answer to vibe design when Google Stitch popularized the category in March 2026. The pitch: stay inside Figma, generate frames and prototypes from prompts, refine in the editor you already use. The reality after running the same brief we used across six AI design tools: fastest tool in the test, weakest output.

We are dMaya, one of the tools Figma Make competes with. The video below is real, produced from the same prompt every other tool got. Watch it before reading the take. The gap between Figma Make's output and the alternatives is visible in 60 seconds.

This is the honest review: where Figma Make wins, where it loses, what to use it for, and when to reach for something else.

Figma Make on the test prompt. Under 60 seconds per prompt. Three screens in two prompts. Output looked first-year-student quality.

What Figma Make actually is

Figma Make is Figma's AI design tool, launched in 2025 and updated through 2026. It runs Claude under the hood (the model is not user-selectable) and lives inside the Figma canvas you already use. Type a prompt, get frames, components, and optionally an interactive prototype back in your existing Figma file.

It was Figma's competitive response to Google Stitch and the broader vibe design category that emerged in March 2026. The strategic logic: keep designers inside Figma rather than letting them leave for prompt-first tools that own the design phase.

Output integrates with your existing components and design system, which is the differentiating selling point. Whether the integration produces work that respects your system in practice is a separate question we will get to.

The hands-on test

Same brief we ran across all six tools in our full prototype tool comparison:

I want to make a SaaS for freelancers where they can do project management and invoicing.

I want to use the nature green as one of the colors. You can plan out the rest of the details and plan features on your own.

I want to start with making Dashboard.

Figma Make handled it in under 60 seconds per prompt. Three screens generated in two prompts (dashboard from the first prompt, time tracking and all-projects screens from a second prompt asking for additional pages). The fastest tool in the entire six-tool test.

The output shipped fast. The output also looked weak.

Figma Make output: freelancer SaaS dashboard

Generic typography. Weak visual hierarchy. Layout decisions that did not commit to the brief. Color choices that ignored the nature-green specification we asked for. The screen works in the "there is something there" sense but does not work in the "a designer made this" sense. The phrase that came up reviewing the output: first-year design student's first project.

Where Figma Make wins

Speed. Fastest tool tested. Under 60 seconds per prompt. For throwaway exploration where speed beats polish, this is real.
Inside Figma. Output lands in your existing Figma file. Normal Figma editing tools work on it. For teams already heavily invested in Figma, the friction floor is the lowest of any AI design tool.
Multi-screen in one shot. Three screens from two prompts is genuinely useful. Many other tools either max out at one screen or need a Pro tier upgrade before they stretch to multiple.
Component-aware. When your existing Figma file has components and styles, Figma Make uses them more often than tools that have no awareness of your design system at all.

Where Figma Make falls short

1. Output quality is the weakest in the category

On the same brief, dMaya (Opus or Sonnet) produced clearly stronger output. Bolt produced clearly stronger output. Even Lovable and v0, which we ranked as average, produced output that looked more committed than Figma Make's. Stitch was the one tool that ranked below Figma Make on output quality, and Stitch is at least free during Labs and honest about being rough.

2. The brief gets ignored

We asked for nature green as a color. The output used a desaturated muddy green that felt arbitrary. We asked for a freelancer SaaS for project management and invoicing. The output looked like a generic dashboard template that could have been generated for any SaaS in any vertical. The brief is not absent from the output, but it does not drive the output.

3. Generic visual signature

Run the same kind of prompt twice and the outputs share a visual signature: gray-on-white cards, system-default-ish typography, safe-corner-radius rounding, low-energy hierarchy. It is the AI-slop aesthetic the rest of the category has been working hard to escape. Figma Make has not yet escaped it.

4. AI credits run out

Figma Make is not free. AI credits are bundled into Figma plans, and serious use exceeds the included allotment fast. For evaluation, plan to upgrade or buy additional credits. The marketing positions Figma Make as included in your Figma subscription. The reality is per-generation cost on top of the seat license.

5. No model picker

Figma Make uses Claude under the hood. You cannot pick which Claude model. You cannot switch to Gemini for cheap exploration. You cannot pick Opus for hero direction. The single-model lock-in means you pay the same rate regardless of the work, which leaves either money or quality on the table compared to tools with explicit model pickers.

Pricing reality

Figma Make is bundled with Figma plan tiers and AI credit allotments. Figma Pro at around $16 per seat per month includes 3,000 AI credits per month. Heavy use requires upgrading the plan or buying additional credits. The pricing model is closer to OpenAI's API than to a flat subscription: you pay per generation.

For comparison: dMaya Starter at $18 per month includes the model picker (Opus, Sonnet, Gemini Flash). Stitch is free during Labs at 350 generations per month. Claude Design is included in Claude Pro at $17 to $20 per month. Figma Make is the most opaque pricing in the category.

What to use instead

Three honest alternatives, depending on what Figma Make is failing at for you.

If Figma Make fails on	Reach for	Why
Output quality	dMaya	Higher output quality on the same brief, model picker per generation, multi-screen consistent, $18/mo Starter.
Cost	Google Stitch	Free during Labs (350/mo cap). Output rough but at least honestly so. Useful for exploration when budget is zero.
Polish on a single screen	Claude Design	Strongest output on a single hero screen. Burns weekly Claude Pro limit fast, so use sparingly. Included in Claude Pro at $17 to $20/mo.

Full same-prompt comparison with all six tools tested side by side, with videos: The Best AI Prototype Tool in 2026.

Want stronger output for the same kind of brief?

dMaya runs Opus, Sonnet, and Gemini Flash through a per-generation model picker on a multi-screen canvas. Plans start at $18/mo.

Start Designing

When Figma Make is the right call

Use Figma Make when

✓ You are already inside Figma and need a rough first pass
✓ You will refine the output manually in Figma's editor
✓ Speed beats polish for the work in front of you
✓ You have AI credits and do not want to switch tools

Skip Figma Make when

✗ Output goes directly to a client without manual refinement
✗ Brand specificity matters in the brief
✗ Multi-screen consistency is non-negotiable
✗ You want to pick the model per generation
✗ Pricing predictability matters more than speed

The fairest read on Figma Make in 2026: it is a 1.0 product. Figma will keep iterating. The current output quality reflects that 1.0 status, and we expect the gap with stronger AI design tools to narrow over time. Right now, for a designer choosing where to spend AI credits today, the better picks live outside Figma.

We will rerun this test when Figma ships a major Make update. The current state is what we documented. Watch the video at the top of the post and judge the output against the alternatives in the full comparison.