Home › Blog › Company logo API

Company logo and brand data API for AI agents

Updated 2026-10-25

This page has affiliate links, each marked "Affiliate link, #ad". We may be paid if you sign up, at no extra cost to you. See our affiliate disclosure.

To get a company's logo from its domain, call a brand data API. Context.dev's Brand API takes a domain, name, email or ticker and returns one profile with logos, colors, company details and socials. Your agent can call it through Context.dev's MCP server, CLI or SDKs. Here are the requests and setup.

How can an AI agent get a company's logo from its domain?

Call a brand data API with the domain. With Context.dev's Brand API, your agent sends POST /brand/retrieve with type by_domain and a domain such as stripe.com, and gets back one profile with logos, colors, company details and social links. The same call works from code, a CLI or an MCP tool. Context.dev's Brand API. (Affiliate link, #ad: we may be paid if you sign up, at no extra cost to you.)

cURL, from Context.dev's Brand docs (key read from an env var):

curl https://api.context.dev/v1/brand/retrieve \
  --request POST \
  --header "Authorization: Bearer $CONTEXT_DEV_API_KEY" \
  --header "Content-Type: application/json" \
  --data '{
    "type": "by_domain",
    "domain": "stripe.com"
  }'

What if you only have a name or an email?

Context.dev accepts other identifiers through the same endpoint: a company name, a work email, a stock ticker, a bank or card transaction descriptor, or an exact page URL. Every lookup returns the same profile shape, so your agent can handle the result the same way whichever input it started from. Store the canonical domain it returns.

What does a brand data API return?

Context.dev's Brand profile includes the domain, title, logos, colors, social links and industry details. Collections can come back empty, so check before use. For full visual tokens such as typography and spacing, its Styleguide endpoint covers a site's design, and Logo Link embeds a logo image directly.

How should your agent pick a logo?

Context.dev's docs suggest choosing a logo by its type, mode and resolution, for example a light or dark version that suits your background. Keep the returned domain next to your input, let people correct uncertain company details, and treat any discovered addresses or contact details as data to verify before relying on them.

Related: data tools in the directory

What can AI agents do with company logos and brand data?

Common jobs are lead enrichment, where an agent adds a logo, industry and socials to a new sign-up from its work email; on-brand documents, where it pulls a client's colors before drafting a deck; and CRM cleanup, where it matches messy company names to one canonical domain.

How do you keep results fresh?

Brand profiles are cached. The maxAgeMs option controls how old a cached profile may be, and setting it to zero forces a refresh. A domain nobody has looked up before can take longer, so give the request a generous deadline or allow partial results, and check the partial flag before treating a profile as complete.

Related: AI agent use cases

How do you connect Context.dev to Claude Code, Cursor or Codex?

Add Context.dev's hosted MCP server at https://mcp.context.dev/mcp and sign in with OAuth, so no API key sits in your config. Claude Code and Codex take one command; Cursor and VS Code take a short JSON entry. Context.dev also offers an agent skill, a CLI and SDKs for TypeScript, Python, Ruby, Go and PHP.

1. Add the server. In a terminal, run the claude mcp add command below to register the context server for your user.
2. Sign in. Start Claude Code, run /mcp, select context and finish sign-in in your browser, or run claude mcp login context.
3. Ask for brand data. Ask your agent for a company's logo and colors by domain, such as stripe.com, and it calls the Brand tool.

Claude Code, from Context.dev's MCP docs:

claude mcp add \
  --transport http \
  --scope user \
  context \
  https://mcp.context.dev/mcp

Codex, from Context.dev's MCP docs:

codex mcp add context --url https://mcp.context.dev/mcp

Cursor (~/.cursor/mcp.json), from Context.dev's MCP docs:

{
  "mcpServers": {
    "context": {
      "url": "https://mcp.context.dev/mcp"
    }
  }
}

VS Code, from Context.dev's MCP docs:

code --add-mcp '{"name":"context","type":"http","url":"https://mcp.context.dev/mcp"}'

Related: MCP context cheat sheet

How do you call it from your own code?

Install the SDK for your language and read the key from the CONTEXT_DEV_API_KEY environment variable, never from the code itself. The TypeScript and Python examples below come from Context.dev's Brand docs. Both call brand.retrieve with type by_domain and print the company's title from the returned profile.

TypeScript, from Context.dev's Brand docs:

import ContextDev from "context.dev";

const client = new ContextDev({
  apiKey: process.env.CONTEXT_DEV_API_KEY,
});

const response = await client.brand.retrieve({
  type: "by_domain",
  domain: "stripe.com",
});

console.log(response.brand?.title);

Python, from Context.dev's Brand docs:

import os
from context.dev import ContextDev

client = ContextDev(api_key=os.environ["CONTEXT_DEV_API_KEY"])

response = client.brand.retrieve(
    type="by_domain",
    domain="stripe.com",
)

print(response.brand.title if response.brand else None)

Related: what an MCP server is

Try Context.dev → Affiliate link, #ad: we may be paid if you sign up, at no extra cost to you.

What else can the same Context.dev key do?

The same account and MCP server also cover web scraping, crawling, search, batches and monitors, so one connection can find a company's site, read its pages and pull its brand profile. For a broader comparison of scraping tools for agents, see our web scraping API guide.

When is a scraper alone enough?

If you only need the text of a page, a scraping API (see Firecrawl MCP server) is enough. A brand API saves work when you need structured fields such as logos, colors and industry for many companies, because it resolves the company first and returns the same shape every time, instead of leaving your agent to parse each site.

Related: agent web search tools trend

See how we run our own agent crew in the playbook.

Building on Grok Bot? Start a crew from the DartzDabz Crew Builder template.

We may be paid by X based on how many people use our Crew Builder template.

See real agent setups you can copy, browse the agent tool directory, or submit a tool; listings are free.