You can now connect your Google Drive with Figma agent → Access Docs, Sheets, and Slides files → Summarize insights and pull in context → Create new files from Figma Available now on Starter+ plans

We’re training AI to get better at the details designers care about. A button can look right but still use the wrong component. An edit can fix one detail and break another. The goal is AI that follows the designer's lead, not just produces a plausible design. That goal shapes how we train and evaluate our first-party editing model for the Figma agent, with our designers helping define what good looks like. Our engineering team shares how we're doing it, and what we're learning along the way. 👇

Excited to share more about the work behind Figma's first-party model: a specialized model for design editing that competes with top models. When we set out to build an agent, we wanted a model that understands design concepts and Figma's tools to help you quickly make edits on the canvas. We started with supervised fine-tuning (SFT) using Figma's native file format, pairing designs with synthetic edit requests. Deterministic validators, visual judges, and a self-healing pipeline helped evaluate and refine our outputs. We've since added reinforcement learning (RL) to our recipe to push for more improvements. Design is subjective, but we can still evaluate structural correctness, visual alignment, and request fulfillment using judges calibrated with expert designers. That includes checking whether an edit preserves everything it should. AI models often create unintended changes like dropped text or mismatched chart labels that we need to catch. Proud of the work of our incredible AI research team. Read more about the model here, and if this work sounds interesting, come join the team!

Shaders and Motion with the Figma agent 00:00 – Intro, Brett McMillin - Design Advocate 01:45 – Automating Motion with the Figma agent 02:45 – Custom Shader Generation 03:00 – Stacking Built-in & Custom Effects 03:28 – Real-time Light Refraction Through Glass

Say hello to iPhone Duo presets wherever you design and prototype in Figma → Properties panel → Prototype settings → Figma Make preview device picker → Code Layers full screen view

Curious how @harvey makes hundreds of tailored demo videos? Through Figma’s MCP, agents work directly on the canvas to: → Pull in visual context from their brand and motion guidelines designed in Figma → Storyboard and animate using Figma Motion, for review and feedback with their team → Generate imagery and video with @figmaweave

In the world of software design and product, we're seeing exponentially more software. But the usage and utilization of apps? Flat. Differentiate with design, a point of view, brand, storytelling, "taste" (and all the buzzwords you can think of)

you'll do your best work in a tool that allows you to be whimsical, too! 🔊🔊🔊 — lol i've lost hours building fun stuff like this with code layers… @figma code layers starting to roll out into open beta TODAY! incredibly excited to see what you'll cook!

Now teams can explore and build code-backed prototypes on the canvas with Code Layers, now in open beta 0:00 Intro 0:43 Starting from code 2:26 Explore code concepts on the canvas 3:57 Edit in both code and design 6:36 Merge ideas together

Today at Config India, we announced new ways of building in Figma → Collab with code on the canvas with Code Layers (open beta) → Shape and share context with Specs (open beta) → Give your agent more design system context with Guidelines (open beta) and Code Sources (closed beta) → Ship websites with Payload (rolling out in the coming weeks)

Announcing 20 new MCP partners at Config India (!!) Amp, ClickUp, Dia, Junie by JetBrains, Linear, Meta Muse, Mistral, Notion, OpenCode, Sarvam Agents, Agentforce Vibes, Amazon Quick, Amplitude Wave, Cline, Delta, Dust, Glass, Gradial, Writer, Zed, Microsoft 365 Copilot