Figma to deploy in-house model and AI credits to defend gross margins

What happened

Figma is reportedly banking on its AI credits and a new in-house model to lift margins, addressing one of the central risks flagged in prior analysis. The master report highlighted that AI inference costs are a key threat to gross margin, which has held near 80% but could compress as AI usage scales. By moving to an in-house model, Figma aims to reduce reliance on expensive third-party AI providers and lower per-inference costs, potentially improving the profitability of AI features. This complements the existing AI credit monetization system that began in March 2026, which has already shown strong usage but with uncertain incremental economics. The market will likely view this as a proactive step, but execution and model quality remain open questions.

Implication

Over the next 6-12 months, successful deployment of a capable in-house model could lift gross margins above current levels and reduce the risk of AI-driven margin compression. This would strengthen the investment case by supporting the base scenario's assumption of stable gross margins near 80%. Conversely, if the in-house model underperforms or the transition proves slower than expected, Figma could face continued margin pressure and competitive disadvantage. That would reinforce the bear scenario of AI-related pricing pressure and margin contraction. Investors should monitor whether the company can demonstrate cost savings from the in-house model in upcoming quarterly filings.

Thesis delta

The introduction of an in-house model is a concrete response to the key margin risk identified in the report. While it does not resolve the uncertainty around AI economics, it signals management's commitment to cost discipline and provides a potential path to margin expansion, modestly increasing conviction in the bull case if execution is successful.