Introducing Parallel Agents - Run multiple /agent queries simultaneously. Read more →
Branding Format v2: Improved Logo Extraction
placeholderEric Ciarla
Feb 06, 2026
Branding Format v2: Improved Logo Extraction image

We've significantly improved logo extraction accuracy for Firecrawl's Branding Format v2, our endpoint for extracting brand identities from websites. It now reliably handles edge cases like logos hidden in background images, buried in unusual HTML structures, or generated by modern site builders like Wix and Framer.

What's new in Branding Format v2

Significantly improved logo detection

Logo extraction is now far more reliable, with fewer false positives and better handling of edge cases like logos embedded in background images. If your tools depend on accurate logo identification, results are now more consistent across the web.

Works with modern site builders

Firecrawl's Branding Format now properly detects logos built with Wix, Framer, and other drag-and-drop platforms that generate complex or non-semantic HTML. Millions of sites are built with these tools — your workflows should work reliably across all of them.

Built for AI agents and developers

Firecrawl's Branding Format goes beyond logos. It captures colors, typography, spacing, and UI components in a structured format designed to power AI agents and applications. One API call gives you the brand data you need to generate on-brand assets, interfaces, and experiences.

How it works

Add branding to your scrape formats to extract a complete brand elements:

from firecrawl import Firecrawl
 
firecrawl = Firecrawl(api_key='fc-YOUR_API_KEY')
 
result = firecrawl.scrape(
    url='https://firecrawl.dev',
    formats=['branding']
)
 
print(result['branding'])

The response includes the site's logo, color palette, typography, spacing scale, and UI component styles — all structured and ready to use.

Use cases

Personalized customer onboarding

Scrape a prospect's website and instantly generate a demo environment that uses their logo, colors, and typography. Instead of manually recreating brand assets for each lead, your onboarding tool does it automatically with a single API call.

Branded asset generation

Build AI tools that generate presentations, landing pages, or marketing materials that match a company's visual identity. Firecrawl's Branding Format gives your agent the structured data it needs (logo URL, hex colors, font families, spacing) to produce assets that actually look on-brand.

On-brand UI components

Generate React components, email templates, or dashboard themes that align with a brand's design system. The structured output maps directly to CSS variables and design tokens, so there's no manual translation step.

Brand monitoring

Track visual identity changes across competitor or partner websites. Detect rebrands, logo swaps, or color palette shifts by comparing Firecrawl's Branding Format snapshots over time.

Give AI agents the brand data they need

Branding Format v2 is available now for all users. Try it today:

  1. Read the Branding Format documentation
  2. Experiment in the Playground
  3. Share feedback on Discord

-- Eric and the Firecrawl team

placeholder
Eric Ciarla @ericciarla
CMO of Firecrawl
About the Author
Eric Ciarla is the cofounder and Chief Marketing Officer (CMO) of Firecrawl. He also worked on Mendable.ai and sold it to companies like Snapchat, Coinbase, and MongoDB. Previously worked at Ford and Fracta as a Data Scientist. Eric also co-founded SideGuide, a tool for learning code within VS Code with 50,000 users.
FOOTER
The easiest way to extract
data from the web
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
Backed by
Y Combinator
LinkedinGithubYouTube
SOC II · Type 2
AICPA
SOC 2
X (Twitter)
Discord