Figma Make Trial

Overview

Figma Make is Figma's AI-driven prompt-to-app tool for creating functional prototypes, web apps, and interactive UI from natural language, existing Figma designs, components, images, and team context. Figma describes Make as a way to bring ideas and existing designs to life through conversation, with an interface centered on AI chat and a working preview where generated code can be inspected and edited (Figma Make help).

The strongest use case is rapid product validation before production engineering begins. Figma introduced Make at Config 2025 as a prompt-to-app and prompt-to-code capability that can transform a written prompt or an existing Figma file into a coded, interactive prototype while preserving design intent and fidelity (Figma Config 2025 recap). Figma later moved Make out of beta into general availability, positioning it as a tool for designers, engineers, and product managers to create high-fidelity artifacts for earlier alignment and more informed handoff conversations (Figma Make general availability).

The reason to classify Figma Make as Trial is that it is promising for design-to-prototype acceleration, but it sits in a sensitive workflow where generated UI can look convincing before it is accessible, secure, performant, maintainable, or production-ready. Teams should pilot it with mature design-system inputs, explicit review gates, and clear boundaries between prototypes, validated product decisions, and production implementation. Treat Figma Make as an exploratory collaboration layer, not as a replacement for design craft, user research, engineering architecture, or production quality assurance.

Adoption Signals

• Figma moved Figma Make and other Figma AI features out of beta in July 2025, making Make available for everyone to try while noting that publishing Figma Make files was still in beta at that time (Figma Make general availability).
• Figma's help documentation says Make is available for Full seats on paid plans, while other seats and plans can try it with constraints; the FAQ further distinguishes features such as file creation, sharing, design attachment, point-and-edit, style context, code editor, code export, and publishing by seat type (Figma Make help, Figma Make FAQs).
• Make supports multiple starting points: users can write text prompts, attach existing Figma designs, attach more than one design for additional guidance, paste frames or components into chat, upload images, and use Community content as input for generated prototypes or web apps (Figma Make help, Figma Make FAQs).
• The product has moved toward more realistic app behavior through a Supabase integration for secret storage, compute, and a Postgres-backed environment, although Figma states that Make sets up basic key-value stores rather than a full SQL database in the provided Supabase Postgres database (Figma Make backend documentation).
• Figma's release notes list Make-specific updates such as custom skills, voice-to-text prompting, question cards, version history, clear context reset, and a Zapier connector for bringing in context from Google Drive, Microsoft Office, Zoom, and other apps (Figma release notes).
• Make integrates with Figma's broader collaboration model: eligible users can collaborate in the same conversation, simultaneously modify generated code, share previews, publish to the public web with a Figma-provided URL, and copy preview snapshots back into Figma Design as editable layers (Figma Make help).

Risks

• Prototype quality can be mistaken for product readiness. Figma Make produces functional prototypes, web apps, and code-backed interactive UI, but teams still need human review for accessibility, responsiveness, edge cases, error states, security, performance, analytics, observability, internationalization, and production architecture (Figma Make help).
• Generated content can create licensing exposure. Figma warns that Make may incorporate third-party internet content such as fonts, code packages, or images, and that users are responsible for ensuring they have appropriate rights before publishing (Figma Make help).
• Publishing is public by default once shared on the web. Figma's FAQ states that published functional prototypes or web apps are available on the public web to anyone with the URL, cannot be access restricted, and do not expose the underlying Make file or conversation to visitors (Figma Make FAQs).
• Seat and plan differences affect rollout. The FAQ says Full seats have the broadest Make access, while Dev, Collab, and View seats are limited mostly to drafts; it also says Starter plans cannot use team libraries for style context and can only publish to the public web if they also publish to Figma Community (Figma Make FAQs).
• Admin settings and AI governance matter. Figma states that organization and enterprise access depends on AI features being enabled, Enterprise admins can manage AI feature availability at organization or workspace level, and content training defaults differ by plan, with Organization and Enterprise content training off by default (Figma AI settings documentation).
• Secrets must not be pasted into prompts. Figma's backend documentation explicitly cautions users to enter secrets and API keys only through the provided Create a secret flow, not as plain text in the AI chat prompt (Figma Make backend documentation).
• Backend support is useful but constrained. Supabase integration supports persistence, authentication-oriented flows, API secrets, and compute, but Make currently sets up basic key-value stores rather than a full SQL database in the connected Postgres environment (Figma Make backend documentation).
• Round-tripping back into design systems has limits. Figma says users can copy a Make preview as layers into Figma Design, but changes to those pasted layers are not applied back to the Make file and the layers are not automatically tied to the design system (Figma Make FAQs).
• Comments and device APIs are incomplete for some workflows. The FAQ states that comments are not currently supported in Figma Make and that Make preview does not currently support microphone and camera access (Figma Make FAQs).

Recommendation

Trial Figma Make for product discovery, clickable concept exploration, interaction prototyping, design-system stress tests, stakeholder alignment, and PM or designer-led proof-of-concepts where speed and fidelity matter more than production robustness. It is especially useful when the team already has strong Figma components, variables, styles, usage guidelines, and review norms, because those inputs give the model a clearer design target and reduce design drift.

Evaluate Figma Make on representative workflows before standardizing it. Good pilots include turning a PRD into a prototype, converting an existing frame into a coded interaction, adding authentication-like flows with Supabase, testing a complex interaction that static Figma prototypes cannot express, and copying Make output back into Figma Design for critique. Measure time-to-first-prototype, design-system adherence, accessibility, responsive behavior, code readability, handoff usefulness, iteration quality, data-handling safety, and whether the prototype improves product decisions.

Adopt governance early. Require teams to label Make outputs as prototypes unless engineering has reviewed them, forbid secrets in prompts, define when publishing is allowed, check licensing for third-party assets, configure AI and content-training settings by workspace, and document which seats can create, share, export, or publish Make files. Move from Trial to Adopt only when teams can repeatedly produce useful prototypes without confusing generated artifacts for production software.

Sources

• Figma Make help
• Figma Make FAQs
• Figma Make general availability
• Figma Config 2025 recap
• Figma Make backend documentation
• Figma AI settings documentation
• Figma release notes