Figma Adopts AI Credits, Expands Partnerships
Figma is launching a usage-based AI credit system in March and is expanding partnerships with model vendors to solidify its role in multi-tool creative workflows. A new update enables a seamless workflow between its design canvas and the Claude Code AI. A recent review also compared the UI design outputs of using Claude 4.6 versus Gemini 3 within Figma's "Make" AI widget, highlighting how model choice affects creative results.
- Figma's AI credit system, starting March 11, 2026, allocates a monthly, non-transferable credit allowance to each user based on their plan, with options for administrators to purchase a shared pool of additional credits. For instance, full seats on a Professional plan might receive 3,000 credits monthly, while Starter plans get 500. - The new workflow with Anthropic's Claude Code allows developers to generate UI in a terminal and then convert the live, running browser state (from localhost, staging, or production) into fully editable Figma layers. This "Code to Canvas" feature aims to reverse the traditional design-first workflow, acknowledging that many interfaces will now originate from AI-generated code. - This integration is built on Figma's Model Context Protocol (MCP), which facilitates the bidirectional workflow, allowing design changes made in Figma to be synchronized back to the codebase via the Figma MCP server. - The debate around AI's role in creativity highlights a shift from AI as a simple tool to a co-creator, raising complex questions about authorship and ownership. Legal frameworks are struggling to keep pace, with most jurisdictions requiring human authorship for copyright protection, often placing purely AI-generated works in the public domain. - In practice, human-AI collaboration is not seen as a replacement for designers but as a partnership that amplifies creativity by handling complex data analysis and automating tedious tasks. This allows designers to focus on user experience, ethics, and translating AI-generated insights into meaningful design decisions. - The rise of multi-tool AI workflows is leading to new, node-based applications like Krea and Freepik Spaces, which allow creatives to chain different models together into cohesive pipelines for tasks like image and video generation. This approach enables practitioners to use the best model for each specific task—such as Gemini for creative writing and a different model for image coherence—and see how changes in one part of the chain affect the final output. - For developers building these creative tools, the landscape of AI-native IDEs is rapidly maturing, with tools like Cursor, Windsurf, and Bolt moving beyond simple code completion to act as "agentic" co-developers that can generate full applications from high-level instructions.
