LinkedIn API pricing in 2026: the free tier, the partner wall, and the cost nobody puts in a table

LinkedIn has an API. It also has no price list. The consumer tier is free and lets you post to your own feed. Everything above that (company pages, analytics, posting for clients, ads, recruiting data) sits behind a partner application with no published fee, no published timeline and, for one product line, no application form at all right now.

So “LinkedIn API pricing” has three honest answers depending on what you’re building: $0, “apply and find out in three to six months,” or “you’ll be paying a third party, and the last big one got sued out of existence.”

Does LinkedIn have an API? Yes, several, and they’re not interchangeable

The Developer Portal lists product families: Consumer, Marketing (which contains the Community Management API and ads), Sales, Talent, Learning, and a newer Verified on LinkedIn API. Each one is applied for separately, each has different access rules, and a token from one doesn’t unlock another.

The Consumer products are the only ones you can turn on yourself. Sign In with LinkedIn (OpenID Connect, so openid, profile, email) gives you name, headline, photo and email. Share on LinkedIn gives you w_member_social, which lets the authenticated member post, comment and react on their own behalf. Published limits are 150 posts per member per day and 100,000 per app. That’s the whole free menu. No company pages, no analytics, no messages, no connections.

For a personal side project that posts to your own profile, that’s enough and it costs nothing.

The Community Management API and the Marketing Developer Platform

The moment you want to touch an organisation page (post to it, read its analytics, moderate comments) you’re in the Community Management API, which lives inside the Marketing Developer Platform. There’s a Development tier for building against test pages and a Standard tier for production, and you request both through the “increasing access” flow in the portal. LinkedIn asks for a product description, a demo, and a privacy posture. There’s no fixed checklist, so you answer whatever the reviewer asks.

LinkedIn does not publish what Standard tier costs. Some blogs print “$699+/month” or “thousands per month”; those are estimates from vendors, not figures LinkedIn has ever released, and you should treat them as guesses. What you can plan around is the timeline: developers report four to eight weeks when the product is polished and the use case is clearly not scraping, three to four months on average, and six-plus months or a rejection when the app looks like it competes with LinkedIn or pulls data at scale.

If your launch depends on Marketing Developer Platform approval, you’re not launching this quarter.

What the docs don’t put in a table

A few things that cost time on LinkedIn specifically.

There’s no scheduling endpoint. The Posts API publishes when you call it. If your product offers “post at 9am Tuesday,” you build the queue and the cron and the retry logic yourself. Developers coming from Meta’s platform expect a scheduled_publish_time field; it isn’t there.

Rate limits are not published. LinkedIn’s own rate-limit page says standard limits aren’t documented and live per app in the Developer Portal’s Analytics tab. The exceptions are the consumer posting limits above and the Community Management Development tier defaults (500 requests per app, 100 per member). Production limits are part of your partner conversation. A 429 is your only signal, and there’s no promise of a Retry-After header.

Marketing endpoints are versioned monthly. Every call goes to api.linkedin.com/rest/ with a LinkedIn-Version: YYYYMM header, and each version lives about 12 months. Consumer endpoints stay on /v2/ without it. Mixing the two hosts is the most common bug in old integrations.

Access tokens last 60 days. Refresh tokens last a year, but programmatic refresh is only available to approved Marketing Developer Platform partners. Outside the partner program, a token that expires means the user reconnects by hand.

And organisation pages are scoped to the member’s role. One admin’s token covers every page where that person is an Administrator, Content Administrator or Direct Sponsored Content Poster. Your customer’s marketing manager needs the right role on the page before your app can post to it, and that’s a conversation your support team will have weekly.

Reading data you don’t own: there is no official door, and the unofficial ones keep closing

Everything above is about writing. If you want to read profiles, companies or posts you don’t control (lead enrichment, recruiting, brand monitoring, competitor tracking), LinkedIn’s official API gives you nothing at scale. The Sales Navigator API is the closest thing and LinkedIn’s Sales API page currently states it isn’t accepting new partners. Sales Navigator seats are still sold at $119.99 a month for Core and $159.99 for Advanced, but a seat is a product licence, not API access. Plenty of developers find that out after buying one.

So the “LinkedIn API pricing” you’ll actually find is third-party data vendors. Coresignal, People Data Labs, Bright Data, Phyllo, Apollo (from $49 a month with a free 10,000-credit tier for contact data), PhantomBuster (from $69) and dozens of RapidAPI wrappers. Enterprise enrichment (ZoomInfo, Clearbit) starts around $15,000 a year.

Price them, but price the risk too. Proxycurl was the default LinkedIn data API for years. LinkedIn sued Nubela, its parent company, in January 2025. The case settled and Proxycurl shut down in July 2025, mid-contract for its customers, with their data slated for deletion. That’s the pattern on LinkedIn: vendors don’t raise prices, they vanish. If your product depends on unofficial profile data, keep the read layer behind an interface you can swap in a week.

The LinkedIn-specific cost: your account

This is what makes LinkedIn different from Instagram or X, and it’s the part most pricing guides skip.

A whole category of LinkedIn “API” and automation tools doesn’t use LinkedIn’s API at all. They log in with your session cookie and drive the site like a browser. Unipile is open about this, and its pricing (from €49 a month for 10 accounts) looks like a bargain next to a partner contract. PhantomBuster works the same way. So do most “LinkedIn automation” Chrome extensions.

LinkedIn’s terms prohibit it, and LinkedIn enforces. Accounts get restricted, sometimes permanently, and there’s no appeal path that reliably works. For a solo creator that’s a nuisance. For an agency, it’s a client’s personal profile, the one attached to their name and their job, getting locked because of a tool you chose. That’s not a line in a price table, but it’s the most expensive thing on this page.

The rule: on LinkedIn, only automate posting through something that publishes via the official API. That means either your own Marketing Developer Platform access, or a scheduler that has it.

If you only need to post, don’t apply

For publishing to personal profiles and company pages, the sensible route for most teams is a scheduler that already holds Community Management access and exposes its own API. LinkedIn reviewed and approved that platform once; you don’t repeat the process.

SchedPilot is one. Every account (personal profile or company page) connects through LinkedIn’s standard OAuth screen inside SchedPilot, so your clients authorise once and you never write the OAuth flow. Then a post is one HTTP request:

curl -X POST https://api.schedpilot.com/developers/v1/post \
  -H "Authorization: Bearer smm_your_key" \
  -H "Content-Type: application/json" \
  -d '{
    "content": "We just shipped webhook retries. Full changelog in the comments.",
    "date": "2026-09-16",
    "time": "08:30"
  }'

Scheduling is handled on the platform (the thing LinkedIn’s API doesn’t do), along with token refresh, versioning headers, media upload and the retry logic for 429s. The same call publishes to X, Instagram, TikTok, Threads, Facebook, YouTube, Pinterest and Bluesky if you tick those accounts too. There’s also an MCP server with six tools (get_accounts, list_posts, create_post, upload_media, delete_post, get_analytics), so Claude, ChatGPT, Cursor or an n8n workflow can draft and schedule LinkedIn posts directly, and anything an agent creates lands in the calendar as pending until a human approves it. Given how LinkedIn treats accounts that post on autopilot, that approval step isn’t a nicety.

Pricing: API access is included on paid plans from $21 a month billed annually (Silver, 5 accounts) and $35 for 20 accounts (Gold). Enterprise at $79 adds 100 accounts, 10 seats and the approval workflow. Against three to six months of partner review plus the engineering to keep 30 client tokens alive, that’s the cheapest LinkedIn API pricing you’ll find that doesn’t put anyone’s account at risk.

What it won’t do: ads data, page-level analytics beyond what the scheduler surfaces, recruiting or Sales Navigator data, and reading profiles you don’t own. If any of those is the product, you need the partner programme or a data vendor, and you should start the application now because the clock is the cost.

Who should do what

Posting to your own profile from a script: use Share on LinkedIn directly. Free, self-serve, 150 posts a day is more than you’ll use.

Building an ads tool, a deep analytics product or an ATS: apply for the partner programme, budget six months, and build something that doesn’t depend on LinkedIn in the meantime. Talk to the partner team before you write code; the review is about your use case, not your implementation.

An agency, a marketing team, or anyone wiring an AI agent to publish to LinkedIn: use a scheduler with official access and an API. $21 to $79 a month, day one, official channel, no restricted accounts.

Lead enrichment or monitoring: pick a data vendor, budget for the day it disappears, and never touch a client’s personal account with a cookie-based tool.

LinkedIn’s price is zero. The bill arrives as months of waiting or a restricted profile. Pay in dollars instead.