Upwork Job Leads Scraper

---

🎯 What this scrapes

Upwork is one of the largest freelance marketplaces, posting thousands of new jobs a day across every skill category. This Actor searches Upwork's own public job-search results by keyword and writes one normalized row per posting: title, posted-age (raw text plus a best-effort ISO-8601 parse), price type and range (hourly or fixed), and experience level. Built as a job-leads / freshness monitor — posted-age is first-class and results are freshness-sortable, so freelancers can re-run the same search on a schedule and catch new postings early instead of running a one-shot bulk scrape.

🔥 What we handle for you

• 🛡️ We clear the challenge once per run — a dedicated browser context absorbs Upwork's access check, then every page in the run reuses it.
• 🌐 We rotate proxies and sessions through Apify Proxy on every block — fresh session ID, fresh exit IP.
• 🔁 We retry with exponential backoff on transport errors — up to 5 attempts per page.
• 🧱 We fail loud, not silent — a hard block surfaces a clear status message instead of a quietly empty dataset.
• 🧊 We keep the dataset clean — Pydantic-validated rows, ISO-8601 timestamps, stable IDs, JSON / CSV / Excel export straight from the Apify Console.
• 💰 Pay-Per-Event pricing — you only pay for results that hit your dataset. No data, no charge.

💡 Use cases

• Fresh-lead monitoring — schedule a run per keyword and catch new Upwork postings before the crowd applies.
• Freelance agency sourcing — track multiple skill keywords at once and feed matches into a shared lead sheet.
• Rate benchmarking — aggregate hourly/fixed price ranges by keyword to gauge going rates.
• Automation pipelines — wire fresh job leads into n8n/Make/Zapier or a Slack/email alert.

⚙️ How to use it

1. Click Try for free at the top of the page.
2. Fill in the input form — most fields have sensible defaults.
3. Click Start. Output streams into the run's dataset.
4. Export from Storage → Dataset as JSON, CSV, or Excel — or fetch via the API.

📥 Input

Field	Type	Required	Default	Notes
searchQuery	string	yes	'python'	Keyword(s) to search Upwork job postings for, mapped to the q= query param. Required.
maxPages	integer	no	5	Page cap for pagination. A depth walk cleared 15-51 pages cleanly on real queries with zero anti-bot re-challenges, but maxPages stays a conservative, operator-configurable hard cap.
maxItems	integer	no	50	Hard cap on dataset rows for this run. Pagination stops as soon as either Max pages or Max items is hit, whichever comes first.
proxyConfiguration	object	no	{"useApifyProxy": true, "apifyProxyGroups": ["RESIDENTIAL"]}	Defaults to RESIDENTIAL — required for this target. A cloud test measured bare datacenter/automatic exits failing outright while RESIDENTIAL cleared 9/10 attempts.

Example input

{
  "searchQuery": "python",
  "maxPages": 3,
  "maxItems": 10
}

📤 Output

Every row is one dataset item.

Field	Type	Notes
job_id	string	Real tile identifier if one is found on the source DOM, else a stable SHA-256 hash (16 hex chars) of tile content. Same…
source_url	['string', 'null']	Canonical job URL parsed from the tile's own link; null only if no per-job link is found on that tile.
title	string	Job posting title, verbatim from the tile.
posted_age_text	string	Raw posted-age text verbatim, e.g. 'Posted 4 weeks ago'.
posted_at	['string', 'null']	Best-effort ISO-8601 parse of posted_age_text; null if the phrasing doesn't match a known relative-date pattern. Raw…
price_type	['string', 'null']	'hourly' or 'fixed', parsed from the tile's price text.
price_min	['number', 'null']	Parsed minimum price from the tile's price range.
price_max	['number', 'null']	Parsed maximum price from the tile's price range.
experience_level	['string', 'null']	Experience level verbatim, e.g. 'Entry Level', 'Expert'.
search_query	string	Echoes the searchQuery input for this row.
page_number	integer	1-indexed source page this row was scraped from.
scraped_at	string	ISO-8601 wall-clock time of row construction.

Example output

{
  "job_id": "a1b2c3d4e5f60718",
  "source_url": null,
  "title": "Python AI Engineer",
  "posted_age_text": "Posted 2 weeks ago",
  "posted_at": "2026-08-31T00:00:00Z",
  "price_type": "fixed",
  "price_min": null,
  "price_max": null,
  "experience_level": "Expert",
  "search_query": "python",
  "page_number": 1,
  "scraped_at": "2026-09-14T12:00:00Z"
}

💰 Pricing

Pay-Per-Event — you pay only when these events fire:

Event	USD	What it is
actor-start	$0.2	One-off warm-up charge per run
result	$0.004	Per dataset item

Example: 1 000 results at the rates above ≈ $4.20. No subscription, no minimum, no card to start — Apify gives every new account $5 of free credit.

🚧 Limitations

v1 covers Upwork's public job-search results only — job detail pages (full description, client history, proposal counts), filters beyond keyword (category, budget, verified-client, experience level), and authenticated views (saved searches, invite-only jobs) are out of scope. A depth walk on real queries cleared 15-51 pages with zero anti-bot re-challenges before hitting each query's own result-set end — but that's one day's evidence, not a permanent guarantee, so maxPages stays a hard, conservative, operator-configurable cap rather than an unlimited crawl. source_url/job_id come from the tile's own listing when present; on the rare tile without one, job_id falls back to a stable content hash so leads still dedupe across scheduled re-runs.

❓ FAQ

Do I need an Upwork account or API key?

No. This Actor scrapes Upwork's own publicly listed job-search results — no login, no API key.

How far back does pagination go?

A depth walk on real queries cleared 15-51 pages cleanly with zero anti-bot re-challenges, ending on the query's own result-set exhaustion rather than a block — but that's one day's evidence, not a guarantee. Max pages defaults conservatively and the Actor reports honestly (via the run's status message) how many pages it actually fetched, whether it stopped early or hit your cap.

Why is source_url sometimes null?

Most tiles carry a real per-job link, which becomes source_url and job_id. On the rare tile without one, job_id falls back to a stable content hash so leads still dedupe across runs.

What if a search matches nothing?

The run succeeds with zero rows and a status message naming the query — a genuine empty search is never reported as a failure.

💬 Your feedback

Spotted a bug, hit a weird edge case, or need a new field? Open an issue on the Actor's Issues tab on Apify Console — we ship fixes weekly and we read every report.

---