Will Figma's AI suite (Make, Sites, Buzz, Draw, Weave) defend the design surface against vibe-coding tools — or will Cursor, v0, Lovable and Bolt embed native design and disintermediate the developer-handoff value pool that drives two-thirds of Figma's MAUs?
Figma Make is Figma's defensive answer to a wave of AI-native prompt-to-app tools that reached commercial scale before Make even launched. Make went GA in late July 2025, powered by Anthropic's Claude with optional Gemini routing, and is currently distributed for free to all paid seats. Adoption inside Figma's enterprise base has been rapid — over 50% of $100K+ accounts now use Make weekly, and roughly 60% of Make files are created by non-designers — but the entire suite has run as a pure cost center, compressing non-GAAP gross margin from 92% to 86% and pushing FY2025 gross margin down to 82%.
The competitive reality is unforgiving. Cursor crossed $1B ARR within a year of crossing $100M, raised at a $29B valuation roughly 2.5x Figma's post-IPO market cap, and reached 18% IDE penetration — the fastest IDE adoption ever measured. Vercel's v0, Lovable and Bolt have each independently scaled past $100M ARR by capturing prompt-to-app workflows that bypass the traditional design-to-dev handoff entirely. Two-thirds of Figma's 13M monthly active users are non-designers, and Figma's own S-1 risk factors flag the possibility of seat erosion in plain language.
| Figma Make | Figma (FIG) | May 2025 beta / Jul 2025 GA | Not disclosed | Public ($21B) |
| Cursor | Anysphere | March 2023 | $1B+ (Nov 2025) | $29.3B (Nov 2025) |
| Lovable | Lovable AB | Late 2024 | $400M (Feb 2026) | $6.6B (Dec 2025) |
| v0 | Vercel | October 2023 | $180M+ (2025) | $9.3B (Sep 2025) |
| Bolt.new | StackBlitz | October 2024 | $40M (Mar 2025) | Private |
| Replit Agent | Replit | September 2024 | ~$265M (2025 est.) | $9B (2025) |
| Bull (--) | Mega-bull (--) | Base (--) | Stagnation (--) | Disruption (--) | |
|---|---|---|---|---|---|
| Per Share | -- | -- | -- | -- | -- |
Quantified gates for shifting probability mass between scenarios
| Bull Prob. | Bear Prob. | Implied Value | Δ from Current |
|---|---|---|---|
| NaN% | NaN% | $NaN/sh | -$NaN |
| NaN% | NaN% | $NaN/sh | -$NaN |
| NaN% | NaN% | $NaN/sh | -$NaN |
| NaN% | NaN% | $NaN/sh | -$NaN |
The S-1 risk factor in Figma's own words
Figma explicitly disclosed in its S-1 that AI 'could alter how consumers and businesses interact with websites and apps... in ways that may result in a reduction in the overall value of interface design, or by otherwise making aspects of our platform obsolete or decreasing the number of designers, developers.' The FY2025 10-K elevated this to a top-line risk factor — 'Competitive developments in AI and our inability to effectively respond to such developments could adversely affect our business.' Management is publicly flagging the disintermediation risk in audited filings.
The bull bridge: MCP + Code Connect
Figma's defensive counter-move is to become the design-context provider TO the AI code editors that could otherwise replace it. The Dev Mode MCP server (GA October 2025) lets Cursor, GitHub Copilot, Windsurf and Claude Code pull authoritative Figma component context into the IDE; the GitHub bidirectional integration shipped March 6, 2026. If this becomes the default AI design-to-code substrate, Figma is leverage on every vibe-coding tool's distribution rather than a casualty of it.
Will the March 2026 AI credit monetization actually offset a meaningful share of inference costs, and how soon will Q1/Q2 2026 disclosures show it?