AnyEvery company logo,
sharp and current.
AnyLogo gives your product the logo, colours and details of any company from a single URL — no logo files to collect, no data contract to sign, and nothing to keep up to date when a company rebrands.
No key and no signup to try it — the dialog calls the public API straight from your browser.
- 0keys to start
- 300+edge locations
- <50mscached response
- 7sizes, 16 to 1024
- 0attribution required
You need a company's logo. Today.
A CRM row. A customer list. A transaction feed. A directory, a portfolio, an investment dashboard. The moment your product shows companies, it needs their logos — and there are only three ways to get them.
-
Hotlink the favicon
It is right there and it is free, so almost everyone starts here. Then you see it: a 16-pixel image stretched across a 64-pixel avatar, soft and blocky next to everything else on the screen. Half of them are a generic globe, and the ones that load at all are whatever the site happened to ship years ago.
-
Collect the files by hand
Someone downloads a few hundred logos, crops them, and commits them to the repo. It looks great for about a year. Then companies rebrand, your customer list grows past the logos you collected, and there is no owner for the folder. Every new company is a manual task nobody wants.
-
Buy an enterprise data contract
The data is good and the logos are current. It also means a sales call, a minimum commitment, and licence terms about where the assets may appear — for a feature that is, in the end, an image in a table row.
You send a domain.
You get back a mark.
There is no upload step, no sync job and no dataset to keep. You put a domain in a URL and an image comes back.
You ask for a domain
Put the domain in an image URL. That is the integration — no SDK, no API key, no account.
We return the sharpest mark
The best asset the company publishes, at the size you asked for. A vector stays a vector, and nothing is ever scaled up past its source.
It is cached at the edge
Stored at the location nearest your user, so repeat requests come back in under 50ms from any of 300+ points of presence.
It stays current
Responses carry the time we last verified them, and anyone can flag a rebrand without a key — so the data corrects itself.
A favicon was never
meant to be big.
A favicon is a 16-pixel image built for a browser tab. There is no detail in it to recover — enlarging it just makes the same sixteen pixels bigger, which is why hotlinked logos always look slightly wrong next to the rest of your interface.
We start from the largest asset a company publishes and scale down, which is the direction that preserves detail. When the source is a vector, it stays a vector and is sharp at any size at all.
Eleven things,
all of them live.
Logos are the part everyone needs first. What makes the rest worth having is that the same domain gets you the company's colours, its details, its symbol and its registration record — from the same place, with the same URL shape.
- 01
A logo that looks right at any size
A favicon is 16 pixels. Stretch it to fill a 64px avatar and it turns to mush. We serve the sharpest asset a company publishes — the vector if there is one — scaled down to the size you asked for and never scaled up past the source.
Seven sizes, from a favicon slot to a hero · Read the reference
- 02
The right shape for the slot
Most companies have two marks: a square icon and a wide wordmark. Drop a wordmark into a round avatar and it gets squeezed into an unreadable smear. Ask for the shape your layout needs and that problem disappears.
Ask for the square mark or the wide one · Read the reference
- 03
Every mark a company has, in one call
When you are building a brand picker or letting a user choose which logo to use, you want the whole inventory rather than one guess. The brand kit returns each variant with its dimensions and content type, flagged scalable when it is a vector.
The whole inventory in one request · Read the reference
- 04
Their brand colours, read from their mark
Tint a table row, a card border or a chart series to match the company it represents. The palette is extracted from the logo itself and ordered by how much of the mark each colour covers, so the first entry is the one to theme with.
Ranked by share of the mark · Read the reference
- 05
The company behind the domain
Name, description, website, social profiles and postal address as plain JSON — no enterprise data contract, no sales call. Fields we cannot verify stay empty instead of being guessed, because blank is more useful than wrong.
Name, description, socials and address · Read the reference
- 06
A company picker with no backendKeyless
Your user types "spot" and expects to pick Spotify. Search turns a name into a domain and a logo URL, and because it is CORS-enabled and needs no key, the whole typeahead can live in browser code with nothing to proxy server-side.
Callable straight from the browser · Read the reference
- 07
Symbols, not just domains
A portfolio screen or a trading feed holds AAPL and BTC, not apple.com. These routes resolve a symbol to its company and redirect to the mark, so a plain img tag is the whole integration.
Stock tickers and crypto symbols alike · Read the reference
- 08
Data that corrects itselfOurs alone
Companies rebrand, and a logo library that cannot hear about it goes stale silently. Every response carries the time we last verified it, and anyone can flag a stale mark without a key at all. We have not found another logo API that lets an unauthenticated caller correct its data.
Anyone can flag a stale mark, without a key · Read the reference
- 09
Your interface never shows a broken image
Nobody has every logo, so the honest question is what happens on a miss. Ours returns a lettered tile coloured from the domain, so the row still looks deliberate. Prefer a collapsed space or your own handling? Ask for a transparent pixel or a real 404.
A lettered tile, empty space, or a clean miss · Read the reference
- 10
Coverage you can check yourself
Every provider advertises a library size, which tells you nothing about whether your domains are in it. We sample random real domains, measure the hit rate, and serve the result from the API — including the segments where we score worst.
Measured rates, served from the API itself · Read the reference
- 11
The official registration recordUK only
Onboarding and compliance flows need more than a logo: the company number, its status, when it was incorporated and the registered office, each linking back to the register it came from. UK companies today, via Companies House.
UK companies today, via Companies House · Read the reference
Anyone can
correct us.
Companies rebrand constantly, and a logo library has no way of knowing. That is why the logos in most products are quietly a year or two out of date — nobody notices, and if they do, there is nothing they can do about it.
So we made reporting a stale mark free and keyless. Anyone who spots one — you, your user, a passer-by — can flag it, and the next request fetches a fresh copy. We have not found another logo API that lets an unauthenticated caller correct its data.
The number most
APIs won't show you.
Every provider advertises how many logos they hold. That number cannot tell you the only thing you care about: whether your companies are in there. So we take random samples of real domains, measure how often we actually return a logo, and publish the result from the API itself — including the segments where we come off worst.
Read the bottom row before the top one. Our coverage is strongest on long-tail domains and weakest on the highest-traffic ones, which is the opposite of the usual pattern — providers with pre-built libraries index popular sites first and thin out down the list. We fetch on demand, so an obscure domain costs us nothing, while a large share of the highest-traffic list is infrastructure: CDN endpoints, ad exchanges and tracking hosts that have no logo because they are not brands. When we do miss, a monogram tile means your interface still renders.
All of that,
from one address.
Every capability on this page is a single request. Point an image at it, fetch it from your server, or call it from browser code — nothing needs a proxy in between, and the size, the shape and the miss behaviour all travel on the same address.
There is no SDK to adopt before any of it works. The docs carry the exact addresses, parameters and responses, with a live request you can run on the page.
Free where it counts.
Logos are the free hook and we mean it: every size, every format, no key, no attribution and no resolution ceiling. What you pay for is contact data and volume.
Free
Logos, variants, kits, colours, search, symbols and fallbacks — keyless at 60 requests a minute. A free key raises that to 600 and adds social profiles and addresses.
Pay as you go
Full company records with emails, phones and legal names. 250k requests included, then $0.05 per 1,000. Cancel whenever.
Credit pack
One million requests that never expire. No subscription — top up when you run out.
Business
Custom volume, a commercial licence and an SLA. Reach us and we will work out what fits.
Full pricing and the questions people ask
Established providers have larger libraries and extras we don't offer yet, and we would rather say so than pretend otherwise. What is different here: the real mark at full quality, no attribution on any plan, no key to start, and data you are free to cache, self-host and keep after you leave.
Put a real logo in your product in a minute.
Nothing to sign up for to try it. Create a free key when you want higher limits and richer company records.