Indeed shut its public API. Here's how to get job data anyway

Does Indeed have an API? What happened to the Publisher API?

No. Indeed has not offered a public job search API since it closed the Publisher Program to new applicants in October 2022, and the program remains closed in 2026 with no open signup and no published rate card (Job Boardly, July 22, 2026). Existing Publisher API keys stopped working in 2023, and Indeed has issued no self-serve key since, paid or free (JobsPipe, July 24, 2026). What remains at docs.indeed.com is a set of employer-side partner APIs, all gated behind an integration request that Indeed must approve before it issues OAuth credentials.

None of those partner APIs returns job listings. The Job Sync API is a GraphQL interface for ATS partners to create, upsert, expire, and check the status of their own postings by ID. The only search-shaped product Indeed still documents for publishers is the Publisher JavaScript Plugin, a hosted widget that renders HTML inside your page rather than returning data. If your goal is to read jobs from Indeed, the official route does not exist, which is why every option below is a scraper.

• Job Sync API: partners push postings into Indeed with qualifications, salary, and benefits; status lookups need the posting ID, and there is no search endpoint
• Indeed Apply: delivers applications from Indeed into a partner ATS; the job must be posted through Job Sync or an XML feed, and Indeed must approve an integration request first
• Disposition Sync API: reports candidate outcomes back to Indeed, no listing retrieval
• Publisher JavaScript Plugin: a hosted search widget that returns rendered HTML, not JSON
• Publisher Program: closed to new publishers since October 2022; existing keys retired in 2023 (sources dated July 2026)

Indeed API alternatives compared: price per 1,000, rating, users

Five scrapers cover the read-Indeed use case, and the table records their live Apify Store numbers as of August 29, 2026. The spread is wide: valig/indeed-jobs-scraper headlines $0.10 per 1,000 jobs (its pricing page shows from $0.07 per 1,000) with a 5.00 rating and 26,654 users, while borderline/indeed-scraper charges $5.00 per 1,000 at 4.63 stars and 22,604 users.

misceres/indeed-scraper is the incumbent with 29,854 users and 2,299 in the last month, yet it carries the lowest rating on the table at 3.76 and costs $3.00 per 1,000. factden/indeed-jobs-scraper sits at $2 per 1,000 (down to $1.20 on higher Apify plans) with a 5.00 rating but only 44 users, so its score rests on very little public evidence. All five bill per event, so you pay for rows delivered rather than compute time.

Datapika's job board scraper is the only multi-source row: $0.005 per job ($5 per 1,000), a 5.0 rating from 3 reviews, 2,471 users with 381 in the last 30 days, and a 100% run success rate. It is not the cheapest Indeed-only feed. It wins on covering Indeed plus LinkedIn, Glassdoor, Google Jobs, ZipRecruiter, Naukri, Bayt, and BDJobs in one deduplicated dataset, with no API key or login.

• Cheapest Indeed-only feed: valig at $0.10 per 1,000 (5.00 rating, 26,654 users, 100% success)
• Most used: misceres at $3.00 per 1,000 (3.76 rating, 29,854 users, 99.7% success)
• Highest-priced single board: borderline at $5.00 per 1,000 (4.63 rating, 22,604 users, 97.8% success)
• Newest: factden at $2 per 1,000 (5.00 rating, 44 users, 100% success)
• Multi-board with dedup and MCP: Datapika at $5 per 1,000 (5.0 rating, 2,471 users, 100% success)

What does it cost to pull 1,000 or 10,000 Indeed jobs?

Pay-per-event pricing keeps the arithmetic simple: multiply rows by the per-row price, and the Apify platform charge is zero for all five actors. For 1,000 Indeed jobs, valig costs $0.10 (1,000 x $0.0001), misceres costs $3.00 (1,000 x $0.003), and Datapika costs $5.00 (1,000 x $0.005). At 10,000 jobs that becomes $1.00, $30.00, and $50.00.

The Datapika number changes meaning when you need more than Indeed. A 10,000-row sweep across Indeed, LinkedIn, and Glassdoor with three single-board actors is three separate runs, three schemas to reconcile, and duplicate postings you pay for more than once. The same sweep through Datapika is one run at $0.005 per deduplicated row, so a job that appears on all three boards is billed once and tagged with every source.

Compare like with like. If Indeed is your only source, valig at $1.00 per 10,000 is 50x cheaper than Datapika and the honest recommendation. If you already run two or more boards, price the whole pipeline, not the Indeed slice.

• 1,000 jobs: valig $0.10, factden $2.00, misceres $3.00, borderline $5.00, Datapika $5.00
• 10,000 jobs: valig $1.00, factden $20.00, misceres $30.00, borderline $50.00, Datapika $50.00
• Datapika bills only deduplicated rows that land in the dataset; a posting seen on three boards costs $0.005 once
• No platform compute charge on any of the five; the per-row price is the whole bill (Apify Store pricing pages, August 29, 2026)

Field coverage: salary, company data, descriptions, deduplication

Every actor on the table returns title, company, location, URL, and a description. The differences are in the structured extras. valig parses salary min and max, benefits such as paid holidays and 401(k), required skills, and job type, and returns descriptions as text and HTML. borderline adds company rating, company logo, emails, latitude and longitude, and urgent-hire and high-volume flags. factden ships a second companies dataset with CEO, founding year, revenue, and social links, and documents date-window sharding past Indeed's roughly 1,000-result depth limit per query.

Datapika returns salary min, max, currency, and interval with optional annual normalization, company size, revenue, rating, review count, industry, and logo, and the full description in markdown or HTML when fetch-description is on. Its distinctive field is the source tag on every row, which only matters because a run can span 8 boards. For Indeed-only extraction, field depth is roughly a wash across the top four; for cross-board work, the source tag and deduplication are the reason to pay more per row.

• Structured salary min and max: valig, borderline, factden, and Datapika all parse it; Datapika adds currency, interval, and annual normalization
• Company enrichment: borderline (rating, emails), factden (separate companies dataset), Datapika (size, revenue, rating, reviews, industry, logo)
• Deep pagination past Indeed's ~1,000-result cap: factden documents it; the others do not advertise it
• Cross-board source tags and deduplication: Datapika only, across 8 boards
• Agent access: Datapika ships a ready MCP link (mcp.apify.com) and needs no API key or login; single-board actors can also run through the Apify API or MCP

Which Indeed API alternative should you choose?

Choose valig when Indeed is your only source and price dominates: at $0.10 per 1,000 with a 5.00 rating and 3,627 monthly users, it is the default for budget feeds. Choose misceres if you want the longest track record; 29,854 users and 99.7% success are real, but budget for the 3.76 rating and $3.00 per 1,000. Choose factden if you must paginate past 1,000 results per query or want the companies dataset, accepting that 44 users means limited public evidence.

Choose Datapika when the question is not how to read Indeed but how to get every relevant job once. Recruiters, hiring-signal pipelines, and agents that search across boards get one schema, one bill at $0.005 per row, and no keys. Skip it if you will only ever query Indeed; at 50x the price of valig for that single board, it is the wrong tool. If you need to post jobs into Indeed rather than read them, none of these applies: file an integration request with Indeed and use the Job Sync API.

• Indeed-only, lowest cost: valig ($0.10 per 1,000)
• Indeed-only, longest track record: misceres ($3.00 per 1,000, 29,854 users)
• Indeed-only, deep pagination and company profiles: factden ($2 per 1,000)
• Indeed plus 7 boards, deduplicated, MCP-ready: Datapika ($5 per 1,000, 5.0 rating)
• Posting jobs into Indeed rather than reading them: request partner access and use the Job Sync API

Questions, answered

Keep reading