Is Your Link in Bio Too Boring? Make One That's Actually Yours with Framer Agent

A link-in-bio page is almost a given if you run a personal brand: one URL for the intro, the work, the socials, and a way to get in touch. Plenty of tools make one quickly. Most of them only let you swap an avatar, a background, and a few button colors. If you want a layout that actually feels like you, that range is thin. The result is usable. If you want it to carry a personal style, I'd start with Framer.

Preview: https://index-linkinbio.framer.website/

Where Framer differs from chat-only AI site tools

AI site tools tend to split into two kinds. One generates from a chat window (v0, Manus). The other is a designer's builder, Framer or Webflow. Framer Agent keeps generation and the canvas in the same place. It can put up the frame; you can still move layers, type, and spacing by hand. Chat-only tools don't give you that surface. When a gap is a few pixels off, you write another prompt, wait, and often miss the proportion you meant. That loop costs time.

	Framer Agent	Chat-only AI tools
How you work	Generate on the canvas, then edit by hand	Text prompts only; hard to touch the layout
Precision	Nudge type and spacing yourself	A few pixels off means another prompt
Pace	Agent drafts the large blocks; you tighten the rest	Small changes mean another round
Best for	Pages that need a brand look and careful type	A quick draft when finish isn't the point

Open Agent and describe the mood in one sentence

The right-hand panel switches between Agent and Style. To generate a new screen, type the mood you want in Agent Chat. You don't have to plan the layout first.

My first prompt was simple:

I want to make a link-in-bio site with a slightly vintage, old-school look, plus a subtle worn texture.

The first pass already surprised me: letter paper, an AIR MAIL stamp in the corner, button borders with a hand-drawn wobble.

First pass from one prompt. The stationery story is complete, though it was not the picture I had in mind.

Interesting, and still not the feeling I was after. A text prompt only carries so many clues. Adjectives rarely lock onto a specific visual.

Add a few reference images

Round two: I found a few images I liked, dropped them in, and asked it to run again in that direction.

What caught me was how closely it picked up type and detail. The hard part of a vintage look is the broken print, uneven ink, a slight bleed, not a yellow filter on top. It handled that well on the headline type.

Second pass after adding a reference. Stationery became newsprint, with headline type that looks a little worn. Closer to what I wanted.

The first pass felt like stamps, colored buttons, sketched borders. The second used newspaper logic: one color, thin rules, hierarchy from type size. Same tool, two brand languages. The difference was a reference image.

Turn repeating pieces into components

Once the style settled, Framer asked whether to gather type and color into a system, and whether repeating blocks (buttons, link lists) should become a component.

Don't skip this. Later, a color change or a type-size tweak, or a pass across several buttons, updates everywhere if a system and components are in place. Without them, you restyle each piece by hand. Where something can be a component, make it one. It saves a lot of time.

Structure first, then content

With the look mostly set, tell it the structure: portrait, bio, a few media links, a footer. Spell that out, then ask it to keep going. Don't re-describe the style for every new block. It will carry the visual language you already locked in.

The default model

I used the default GPT 5.6 in the Framer Agent panel, listed as GPT 5.6 Terra, OpenAI's balanced option.

I didn't switch to a more expensive model. Every screen in this note came from the default.

The default is enough to start. A higher-tier model will usually get the details closer.

Three ways to get finer results

A few things I noticed in practice:

Section by section
Asking for a whole page at once tends to come back flat, with fewer details. One section, one sentence, one header or banner at a time gets closer to what you actually want.
Drop photos into the chat
You don't have to upload a photo, then place it by hand. Put it in AI Chat, say where it goes or which image to replace, and it finishes the placement.
Let the Agent edit the image
Ask to make a photo black and white, a little brighter, or higher contrast, and it can do that in Framer's image tools. You don't have to jump out to another editor.

The finished site

After the style, the system, and the structure were in place, the rest was detail work on each block. There isn't much new to say here. Pick a section, say what you want, look, adjust.

The final version uses a beige interactive grid, social icons, a YouTube block, featured products, a desk setup, Tech Stack, an Instagram gallery, and a subscribe flow. You're welcome to browse the live site. I've also left it open to remix for free.

Remix this link in bio

Beige interactive grid, social icons, YouTube, featured products, desk setup, Tech Stack, Instagram, and subscribe. The live site has the full page.

If you want a link in bio with more of a memory to it, you don't need the whole structure first. Start with one sentence about the mood, then a few rounds of edits. The free plan's 500 credits are enough to feel how it works.

Start a Framer link in bio