Fresh LinkedIn data, in bulk, through one API.

Your first call

Everything is a batch, including a batch of one. You submit and you read its results - one shape, from a single profile to fifty thousand, with no ceiling to discover.

curl -X POST https://api.easydata.win/v1/profiles/enrich \
  -H "X-API-Key: $API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"targets": ["https://linkedin.com/in/satyanadella"]}'

{
  "data": {
    "batch_id": "afec9131-080b-4b96-b3bb-726249e4bbf8",
    "operation": "profiles.enrich",
    "status": "queued",
    "total": 1,
    "recommended_poll_ms": 2000
  },
  "meta": { "requestId": "req_8f2ka91x" }
}

Then read it, as many times as you like while it fills in:

curl "https://api.easydata.win/v1/batches/afec9131-080b-4b96-b3bb-726249e4bbf8/results" \
  -H "X-API-Key: $API_KEY"

When something is waiting on the answer, add /sync to the same path and send one target instead of a targets array: the record comes back in the response instead of from a poll, for twice the credits. Synchronous requests is the whole of it.

Authentication

One header, on every request.

X-API-Key: pk_live_...

Base URL: https://api.easydata.win/v1

What a credit is

What a credit costs

A credit is a unit of upstream work. What one costs in money falls with the size of the order - $3 per 1,000 credits at 10,000, down to $1 per 1,000 at 1,000,000.

The full ladder, with worked examples

All operations

Guides

• QuickstartYour first batch, submitted and read, in two calls.
• SDKs and toolsOfficial clients for Python, TypeScript and Go, and an MCP server for agents.
• AuthenticationOne header, on every request - plus scopes and expiry, so a leaked key is bounded.
• What a credit isOne rule, and every price on this site derived from it.
• The batch modelEverything is a batch, including a batch of one.
• Streaming resultsThe same cursor over a held connection, for a client that cannot receive a webhook.
• Synchronous requestsOne entity, answered in the response, at twice the price.
• ErrorsOne envelope, a fixed vocabulary, and one promise about billing.
• Rate limitsFour ceilings, the headers that report them, and how to back off.
• Finding email addressesSet find_emails and every person you scrape comes back with a work email address beside them, for a flat +4.
• WebhooksSix events - started, each row, completed, failed, and two about credits - signed two ways and retried with backoff.

Schemas

One page per record, field by field - what it is, and what it is not.

• ProfileA person, as profiles/enrich returns them. The sample below is the fullest form - what profiles/enrich/detailed returns with every section ticked - because this page lists what the record can carry; the two counts, six of the position fields, skill_details, interests and recommendations each cost a section and are absent without one.
• MonthDayMonth and day of birth, nothing else. LinkedIn never returns a birth year, so this is the whole fact rather than a partial date - contrast with PartialDate, which does carry a year.
• CertificationOne certification or licence the member added.
• LanguageOne language the member listed, and how well.
• SkillOne skill as the profile's skills SECTION carries it, which is the flat skills list plus the two things that make it worth buying: who endorsed it and what the member attached it to.
• InterestOne thing a member follows. All five categories render as the same row on LinkedIn, which is why category is the field that tells them apart.
• RecommendationOne written recommendation. Which way round it runs is the fact about it, so direction and name are read together: name is always the OTHER person.
• ProfileStubA person as a Sales Navigator search card describes them.
• CompanyA company, as companies/enrich returns it.
• CompanyStubA company as briefly as we know it: a search card, or a company another company names.
• EmailFindingsA work email lookup for one person. An object rather than a bare string, because an address is the end of a process that can succeed, half-succeed, honestly fail, or never have started.
• EmailCandidateOne address that was generated and put to the mail server.
• PositionOne role at one company.
• EducationOne school.
• PartialDateA date LinkedIn only partly knows.
• AddressA company's headquarters, component by component.
• LocationOne of a company's offices, with the group LinkedIn files it under.
• SearchPageThe data of one Sales Navigator result entry: one upstream page.
• PaginationWhere one page of a search stands.
• BatchA submission and its progress.
• BatchResultEntryOne delivered result.
• AccountYour allowance, credit balance, limits and callback settings.
• UsageA rollup by day and operation.
• ErrorThe failure envelope. Every failure, without exception.
• MetaThe second half of every success envelope. camelCase, where data is snake_case.
• ProfileActivityThe three activity kinds in one record: posts and reposts, comments and reactions. profiles.posts, profiles.comments and profiles.reactions each return this same shape, with the other two arrays absent.
• ActivityPartialWhich kinds were requested and could not be read.
• PostA post or repost, as profiles.posts and profiles.activity return it. posts.enrich returns this same record by URL or bare urn, with comments included.
• PostCommentA comment on a post, with its author.
• PostAuthorA feed actor is these four fields.
• PostMediaOne image, video or document attached to a post.
• PostArticleThe link preview on a post that shares an external article.
• PostMentionOne person or company @mentioned in a post.
• ReactionCountOne reaction type and its tally, from the breakdown behind reaction_count.
• ActivityCommentOne comment and the post it was left on.
• ActivityReactionOne reaction and the post it was left on.
• ActivityTargetPostThe post a comment or reaction was left on, as a stub.

Machine-readable

• OpenAPI 3.1The spec every reference page is generated from.
• llms.txtFor AI assistants and editors.