Best Indeed scrapers: $0.10 to $6.00 per 1,000 jobs, compared

Why not use the official Indeed API?

Because there is no self-serve Indeed API for reading job postings. Indeed rolled back the Publisher Program to new applicants in October 2022 and retired the Publisher API in 2023 without a self-serve replacement (Job Boardly, Indeed affiliate program status, 2026; JobsPipe, Indeed API key explainer). The misceres scraper's own README, live on August 29, 2026, states that the Publisher Jobs API (Get Job and Job Search) has been deprecated.

What Indeed still offers is employer-side. The Job Sync API publishes and updates an employer's own postings, Indeed Apply receives applications, and Sponsored Jobs manages campaigns, per docs.indeed.com. Job Sync can return a posting's status and details, but only to the ATS that created it under an OAuth token. None of these let a third party search or read other employers' listings, and access is a partner review rather than a signup form. Enterprise data licensing exists but is NDA-gated with no published pricing. That leaves scrapers as the only practical way for a job board, researcher, or AI agent to read Indeed postings. The Apify Store lists dozens of Indeed actors; this page ranks the six most visible Indeed-only ones plus Datapika.

• October 2022: Publisher Program rolled back to new applicants and still closed in 2026 (Job Boardly)
• 2023: Publisher API retired, existing keys stopped working, no self-serve equivalent since (JobsPipe)
• Remaining Indeed APIs move data toward Indeed, not out of it: Job Sync, Indeed Apply, Sponsored Jobs (docs.indeed.com)
• Data licensing is enterprise-only and NDA-gated with no public price
• Every actor below reads public listings without an Indeed account or API key

How the 7 Indeed scrapers compare on price, rating, and users

The table is a snapshot taken August 29, 2026 from the store pages and the public Apify store API. Three facts stand out. First, the advertised "from" price is often a Gold-plan price: misceres shows from $3.00 per 1,000 but bills $6.00 on the free plan, and automation-lab shows from $1.80 but bills $3.00 per 1,000 plus $0.005 per run below Gold. Second, price and popularity are inversely related: misceres leads on users (29,854) with a 3.76 rating from 64 reviews, while valig charges $0.10 per 1,000, 60 times less at base tier, and holds 5.0 from 17 reviews with 26,654 users. Third, borderline costs $5.00 flat and keeps a 4.63 rating from 37 reviews.

Datapika bills $5.00 per 1,000 and does not win on price. On raw 30-day success it is last among the working actors at 90%, because aborted and timed-out runs count against it. Its row wins on scope: one input covers Indeed, LinkedIn, Glassdoor, ZipRecruiter, Naukri, and Bayt by default (Google Jobs and BDJobs are selectable but currently return nothing), deduplicates across them, and returns salary and company fields where the board exposes them. Its README documents MCP access for agents.

• Cheapest: valig at $0.10 per 1,000 on the free plan, $0.07 on Gold and above, 99.7% raw success over the last 30 days
• Most used: misceres at 29,854 users and 2,299 monthly, but the lowest rating in the set at 3.76
• Highest-rated with real volume: valig (5.0, 17 reviews) and borderline (4.63, 37 reviews)
• Most expensive at base tier: misceres at $6.00 per 1,000; borderline and Datapika are $5.00 flat
• Avoid for now: smorgi_apps, offline until August 30, 2026 and then rising to about $15.00 per 1,000 per its README

What 1,000 and 10,000 Indeed jobs actually cost

All seven actors bill per delivered row, so the arithmetic is linear. Using free-plan prices from the store pages and API on August 29, 2026: valig at $0.10 per 1,000 means 1,000 jobs cost $0.10 and 10,000 cost $1.00, plus $0.001 per run per GB of memory. Misceres at $6.00 per 1,000 means $6.00 for 1,000 and $60.00 for 10,000; a Gold plan drops that to $3.00 and $30.00. Automation-lab at $3.00 per 1,000 plus $0.005 per run means $3.005 for 1,000 in one run and $30.005 for 10,000. Factden at $2.00 per 1,000 means $2.00 and $20.00. Borderline and Datapika at $5.00 per 1,000 mean $5.00 and $50.00.

The Datapika number needs one adjustment. Its rows are deduplicated across boards, so a posting on Indeed, LinkedIn, and Glassdoor is delivered and billed once, tagged with every source. Buying 10,000 rows from an Indeed scraper and 10,000 from a LinkedIn scraper means paying twice for each cross-posted job and merging the files yourself. For an Indeed-only feed, valig's $1.00 per 10,000 is unbeatable. For a multi-board dataset with normalized salaries and company size, revenue, rating, and industry, Datapika's $50 per 10,000 replaces several scrapes and a dedup step.

• valig: $0.10 for 1,000 jobs, $1.00 for 10,000, plus $0.001 per run per GB
• misceres: $6.00 for 1,000, $60.00 for 10,000 on the free plan; $3.00 and $30.00 on Gold
• automation-lab: $3.005 for 1,000 in a single run, $30.005 for 10,000; $1.80 per 1,000 on Gold
• factden: $2.00 for 1,000, $20.00 for 10,000; $1.20 per 1,000 on Gold and above
• borderline and Datapika: $5.00 for 1,000, $50.00 for 10,000; Datapika rows are deduplicated across boards
• New Apify accounts get $5 in free monthly credit, which factden's page notes is roughly 2,500 jobs at $2.00 per 1,000

Field coverage: salary, company data, and descriptions

Price per 1,000 is only half the comparison; the rows differ in shape. Datapika returns on each row the title, company, location, job URL, a direct apply URL when available, and the full description in markdown or HTML, plus salary minimum, maximum, currency, and interval with an enforceAnnualSalary option that converts wages to yearly equivalents, plus company size, revenue, rating, industry, and logo where the board exposes them. Filters cover remote only, job type, posted within N hours, distance radius, and easy apply, and a country setting switches Indeed between markets such as the US, UK, or India.

Among the Indeed-only actors, enrichment varies. Factden's page highlights a native salary-range filter, every result page past the roughly 1,000-result ceiling, and free company profiles (rating, size, industry, CEO, founding year, website) in a separate dataset, with no login. Automation-lab's page states no Indeed account is needed. Misceres remains the reference for a plain listing schema; valig emphasizes price and reliability over field depth. Check whether pay comes back as structured fields or a raw string; Datapika does the former and normalizes hourly, monthly, and annual figures so a $45 per hour contract sorts next to a $95,000 salaried role.

• Datapika: structured salary (min, max, currency, interval) with optional annual normalization
• Datapika: company size, revenue, rating, industry, logo, and description alongside each posting where available
• Datapika: full description in markdown or HTML, direct apply URL when available, remote flag, job type
• factden: salary-range filter, all result pages, separate free company-profiles dataset, no login (store page, August 29, 2026)
• automation-lab: no Indeed account or API key required; pricing revised August 25, 2026 to $3.00 per 1,000 plus $0.005 per run on the free plan
• All seven actors read public listings only; none returns applicant or employer private data

Which Indeed scraper should you choose?

Match the tool to the job. If you need raw Indeed listings at the lowest cost and will enrich them yourself, choose valig: $0.10 per 1,000, a 5.0 rating, and 99.7% of 214,408 runs succeeding in the last 30 days (August 29, 2026). If you want the longest track record and largest community, misceres has 29,854 users, but budget $6.00 per 1,000 on the free plan ($3.00 on Gold) and read the 64 reviews behind its 3.76 rating first. Borderline is the middle ground for teams that value a 4.63 rating from 37 reviews over price; its 96.5% raw success rate is the lowest of the three big actors.

Choose Datapika when Indeed is one input among several. Salary research, hiring-signal monitoring, and feeds that must not double-count cross-posted roles benefit from one deduplicated dataset instead of separate Indeed, LinkedIn, and Glassdoor exports, and its README documents MCP access for AI agents. Its 90% raw success rate counts user-aborted and timed-out runs. Skip smorgi_apps, offline until August 30, 2026 with a rise to about $15 per 1,000 announced in its README. Factden and automation-lab are promising but lightly tested: near-perfect rates on only 996 and 873 runs.

• Cheapest Indeed-only feed: valig at $0.10 per 1,000
• Largest user base and longest history: misceres at $6.00 per 1,000 on the free plan, 3.76 rating
• Best reviews among the $5-per-1,000 Indeed-only actors: borderline at 4.63 from 37 reviews
• Multi-board dedup, salary normalization, company enrichment, and documented MCP access: Datapika at $5.00 per 1,000
• Salary-range filtering on Indeed alone: factden at $2.00 per 1,000
• Not yet: smorgi_apps, offline until August 30, 2026 and moving to about $15 per 1,000

Questions, answered

Keep reading