Magnific introduces Designer node in Spaces for multi-page layouts

• Magnific's new Designer node turns Spaces into an in-canvas layout tool, and magnific's launch post says it can generate images and text, place them directly in a design, and keep everything inside the same workflow.
• The official Designer node docs add the useful detail: designs can be multi-page, use placeholders for AI-generated inputs, and send each finished page downstream as its own image.
• magnific's Auto Layers post expands the editing side of the stack by splitting an imported image into editable text, subject, and background layers across Designer, Spaces, and MCP.
• Magnific's MCP rollout is already live on paid plans, according to magnific's paid-plan note, and the MCP page positions it as a way to run generation, editing, resize, and saved workflows from Claude, ChatGPT, Cursor, Codex, Gemini, and other clients.
• One caveat surfaced in magnific's credit reply: MCP requests always consume credits, even for users who may have unlimited generations elsewhere in the product.

You can jump from the Designer node blog post to the full Designer node docs, then over to Magnific MCP, where the company lists Claude Code, Cursor, VS Code, Gemini, and ChatGPT as supported surfaces. The most interesting connective tissue is that magnific's Auto Layers demo is not framed as a standalone edit tool, it is framed as something that now works across Designer, Spaces, and MCP.

Magnific is pitching Designer as a full editor inside Spaces, not a separate export destination. magnific's launch post lists the core promise in one line: generate images and text, drop them into the layout, then keep building on a multi-page canvas with custom fonts and effects.

The official Designer node docs fill in the mechanics:

• Multi-page documents, with each page exported separately.
• Text, images, shapes, and effects in the same canvas.
• Placeholders that pull inputs from other Space nodes.
• Finished pages that flow out to downstream nodes as images.

That last bit is the real workflow shift. The blog post describes Designer as the place where layout, AI-generated content, and collaborative workflows meet on one canvas, which is a much more specific claim than just "design editor in browser."

Auto Layers is Magnific's answer to the usual AI-image editing pain point: one typo or one bad subject placement forcing a full rebuild. In magnific's demo, the company says a dropped image can be broken into three editable parts:

• Text
• Subject
• Background

The MCP product page already listed layer editing as part of the chat-driven workflow, but magnific's announcement makes the scope explicit by saying Auto Layers now works in Designer, Spaces, and Magnific MCP. That means the same layer-aware edit model is showing up both in the visual canvas and in agent-driven clients.

Magnific is not keeping this workflow inside its own app. magnific's integration post says MCP works with Claude, Slack, ChatGPT, Cursor, Codex, and more, while magnific's Gemini reply adds Gemini to the supported list.

The official MCP documentation adds two useful implementation details:

• It is a remote MCP server, so users do not need to run a local bridge.
• It runs on the same Magnific account and credit balance as the main product.

Two reply tweets sharpen the practical limits. magnific's existing Spaces reply says existing Spaces can be used through MCP, and magnific's credit reply says unlimited generations do not carry over there, because every MCP request consumes credits.