Upwork Jobs Scraper: Pay, Clients & Job Alerts, No Login

Use this Upwork jobs scraper to get public Upwork job posts by search phrase, search link, job link or job ID. Find clients, track freelance demand or get new job alerts. Each row has pay, skills, applicants and public client data such as country, spend and rating. No Upwork account or cookies needed.

Start with queries: ["web scraping"] and maxItems: 10. The new Input form has this example filled in. Pricing is $0.001 per run start + $0.0001 per source request + $0.90 per 1,000 saved jobs on the Free plan (down to $0.63 on Gold). Check RUN_SUMMARY to see why a run saved fewer jobs than you asked for.

Table of contents

• What it does
• How it works
• Quick start
• Use cases
• Input
• Output
• Pricing
• Limits
• API, export and integrations
• Troubleshooting
• FAQ
• Related Actors
• Support

What it does

The Actor reads public job posts from Upwork. You can search by phrase, paste a search link from Upwork, or read known jobs by link or ID. You do not need an Upwork account, cookies or your own proxy.

Each job is one Dataset row. The row has the title, description, budget or hourly rate, payment type, experience level, skills, duration, workload and post time. With public job details on, it also has the client country, spend, rating, reviews, hires, applicants and freelancer requirements.

Use filters to keep only the jobs you want. Use Only new jobs on a schedule to get new jobs after each run.

Good fit

• Find new clients and leads for a freelance team or agency.
• Track demand, pay and skills in a job category.
• Get an alert when a new job matches your search.
• Feed job rows to a sheet, database, Slack, Notion or an AI agent.

This Actor does not log in, read private or invite-only jobs, send proposals or contact clients. It does not read freelancer profiles. It does not get client names, emails or phone numbers.

👉 Try it now: open the input form

How it works

1. You give search phrases, Upwork links or job IDs. Bad input stops the run before any source request.
2. The Actor reads public search pages. It then reads public details for the jobs it found, unless you turn includeDetails off. It slows down when Upwork asks it to.
3. It applies your filters and saves one row per unique job. It stops at maxItems or at the end of the search.
4. It writes a RUN_SUMMARY report with counts and the reason the run stopped. It can also post a short summary to Slack or Notion.

👉 Try it now: open the input form

Quick start

1. Open the Input tab. Keep web scraping in Search phrases, or type your own phrase (box 1).

2. Keep 10 in Maximum results for a first test (box 2). Filters are optional, see Input.

3. Click Start.

4. When the run ends, open the Output tab. Each row is one job (box 4). Check the job title, budget, client spend and applicants.

   

Expected result: up to 10 job rows from the newest public jobs for your phrase. Live jobs change all the time, so your rows will differ. An API call must give its own search phrase, link or job ID. An empty API input stops with EMPTY_INPUT.

👉 Try it now: open the input form

Use cases

Buyer job	Input path	Output entity	Key fields	Next integration
Find clients who pay well	queries, minClientTotalSpent, paymentVerified: true	One job per row	title, clientCountry, clientTotalSpent, clientRating, url	CRM or a sheet
Get new job alerts	queries, onlyNew: true, Apify Schedule	Only jobs not saved before	title, budget, hourlyRange, postedAt, url	Slack or Notion with deliveryConnector
Track market demand and pay	queries or categories, maxItems: 1000	One job per row	skills, hourlyBudgetMin, hourlyBudgetMax, budget, category	Sheet or BI tool
Copy a search you made on Upwork	startUrls with an Upwork search link	One job per row	title, jobType, experienceLevel, postedAt, url	Dataset API
Check known jobs	jobIds or job links in startUrls	One row per public job	jobStatus, totalApplicants, hired, clientLastActivity, detailsStatus	Your own job tracker

👉 Try it now: open the input form

Input

Give at least one of queries, query, startUrls, rawUrl or jobIds. An empty input stops with EMPTY_INPUT. All other fields are optional. Filled filters all apply together (AND). Duplicate jobs are saved once. The largest run we tested saved 6,000 jobs; maxItems accepts more, but larger runs are not proven.

Input fields

Field	Plain name	Type	Required	Default / prefill	Allowed values	Effect	Example	Recommendation
queries	Search phrases	array	One of the source fields	Prefill: ["web scraping"]	Any	Search each phrase. Results from all searches share the result limit. Leave blank when you use URLs or job IDs.	["web scraping"]	Start here.
startUrls	Search or job URLs	array	One of the source fields	None	Public upwork.com search and job links	Public Upwork search or job links. Search filters in a link take the place of the same input filters.	[{"url":"https://www.upwork.com/nx/search/jobs/?q=python"}]	Use to copy a search from Upwork.
jobIds	Job IDs	array	One of the source fields	None	Numeric ID or ID that starts with ~02	Read public details for these job IDs. Use the numeric ID or the ID that starts with ~02.	["2117261165716622937"]	Use for known jobs.
maxItems	Maximum results	integer	No	Default and prefill: 10	1–1,000,000	Stop after saving this many matching jobs across all inputs. The source may have fewer jobs. Pages read before this limit still count as source work.	100	Use 10 for a first test.
includeDetails	Read public job details	boolean	No	Default: true	true, false	Include the public client panel, job activity and requirements. A job that is visible only to members keeps its search fields.	true	Keep on. Turn off for a cheaper list with search fields only.
mode	Run mode	string	No	Default: search	search, details, monitor	Search finds jobs. Details reads job links or IDs. Monitor keeps a history and returns new or changed jobs.	"monitor"	Keep search unless you read known jobs or need a monitor.
jobType	Payment types	array	No	None	hourly, fixed, weekly_retainer	Keep hourly, fixed-price or weekly retainer jobs.	["hourly"]	Use it, as it narrows the search itself.
experienceLevel	Experience levels	array	No	None	entry, intermediate, expert	Keep the selected experience levels.	["expert"]	Use it, as it narrows the search itself.
categories	Categories	array	No	None	12 Upwork categories	Upwork category IDs. Choose a category by name. Leave blank for all categories.	["531770282580668418"]	Pick by name in the form.
minBudget	Minimum fixed budget	number	No	None	0 or more	Keep jobs with a known fixed budget at or above this amount.	500	Use for lead lists.
maxBudget	Maximum fixed budget	number	No	None	0 or more	Keep jobs with a known fixed budget at or below this amount.	5000	Optional.
minHourlyRate	Minimum hourly rate	number	No	None	0 or more	Keep jobs whose known minimum hourly rate is at or above this amount.	30	Use for lead lists.
maxHourlyRate	Maximum hourly rate	number	No	None	0 or more	Keep jobs whose known maximum hourly rate is at or below this amount.	100	Optional.
workload	Weekly workload	array	No	None	part_time, full_time, as_needed	Filter by the weekly time needed.	["full_time"]	Optional.
duration	Project duration	array	No	None	week, month, semester, ongoing	Filter by the source project duration group.	["month"]	Optional.
contractToHire	Contract-to-hire only	boolean	No	Off	true, false	Keep jobs marked as contract-to-hire. Needs public job details.	true	Optional.
includeKeywords	Required words	array	No	None	Any	Each word or phrase must appear in the title, description or skills.	["python"]	Use for must-have words.
excludeKeywords	Excluded words	array	No	None	Any	Skip a job if any word or phrase appears in its title, description or skills.	["wordpress"]	Use to remove noise.
requiredSkills	Required skills	array	No	None	Any	Each skill name must be in the job skills. Match ignores letter case.	["Python"]	Use exact skill names.
titleContains	Words in title	string	No	None	Any	Keep jobs whose title contains this phrase. Match ignores letter case.	"scraper"	Optional.
maxAgeMinutes	Maximum job age in minutes	number	No	None	0 or more	Keep jobs published within this many minutes.	1440	Use for fresh jobs.
maxJobAge	Job age	object	No	None	Units: minutes, hours, days, weeks	Alternative age filter with value and unit. Units: minutes, hours, days or weeks.	{"value":2,"unit":"days"}	Same as maxAgeMinutes, but easier to read.
fromDate	Published from	string	No	None	YYYY-MM-DD or ISO time ending in Z	Keep jobs published on or after this UTC date or time. Use YYYY-MM-DD or an ISO time ending in Z.	"2026-11-01"	Optional.
toDate	Published through	string	No	None	YYYY-MM-DD or ISO time ending in Z	Keep jobs published through this UTC date or time. A date without a time includes the full day.	"2026-11-04"	Optional.
proposals	Proposal ranges	array	No	None	0-4, 5-9, 10-14, 15-19, 20-49, 50-	Keep jobs with an exact applicant count inside these ranges. Needs public job details.	["0-4"]	Use 0-4 for low competition.
maxApplicants	Maximum applicants	integer	No	None	0 or more	Keep jobs with a known applicant count at or below this value. Needs public job details.	10	Use for low competition.
paymentVerified	Payment verified only	boolean	No	Off	true, false	Keep clients with a verified payment method. Needs public job details.	true	Use for lead lists.
minClientTotalSpent	Minimum client spend	number	No	None	0 or more	Keep clients with a known public lifetime spend at or above this amount. Needs public job details.	10000	Use to find proven buyers.
minClientRating	Minimum client rating	number	No	None	0–5	Keep clients with a known public rating at or above this value. Needs public job details.	4.5	Removes clients with no reviews.
minClientReviewCount	Minimum client reviews	integer	No	None	0 or more	Keep clients with at least this many public reviews. Needs public job details.	5	Optional.
clientHistory	Client hiring history	array	No	None	0, 1-9, 10-	Keep clients with no hires, 1–9 hires or 10+ hires.	["10-"]	Use 10- for repeat buyers.
clientCountries	Client countries	array	No	None	Any	Client country names, such as United States. Needs public job details.	["United States"]	Use for a market or language.
excludeLocations	Exclude client countries	array	No	None	Any	Skip jobs from these client countries. Needs public job details.	["India"]	Use to drop markets.
timezones	Client timezones	array	No	None	Any	Upwork timezone filter values from a search link.	Value from a link	Copy from a link.
onlyNew	Only new jobs	boolean	No	Off	true, false	Keep a history across runs and return jobs not saved before. Use Return changed jobs to include edits.	true	Use on a schedule.
emitChanged	Return changed jobs	boolean	No	Off	true, false	With Only new jobs, also return jobs whose public fields changed.	true	Use to track edits.
stateName	History name	string	No	None. History uses upwork-jobs-scraper-state	Letters, digits, hyphens; up to 51 characters	Use the same history name on later runs. Use a different name for a separate monitor. Only needed for Only new jobs.	"python-jobs"	Use one name per monitor.
resetState	Reset job history	boolean	No	Off	true, false	Ignore the previous history for this run. New history contains only jobs saved by this run.	true	Use once, then turn off.
skipReposts	Skip repeated job text	boolean	No	Off	true, false	Keep a history and skip a new job ID when its cleaned title and description match a job saved before.	true	Use for alerts.
deliveryConnector	Send results to (optional)	string	No	None	Any	Your Slack, Notion or other MCP connector. After the run the Actor posts a short summary there. Connect services once in Apify Console → MCP connectors.	Your connector	Optional.
deliveryTarget	Notion page or channel (optional)	string	No	None	Any	Where to post the summary: a Notion page link (each run is added on top of that page) or a Slack channel ID. Leave empty to let the connector decide.	Slack channel ID	Optional.
descriptionFormat	Description format	string	No	Default: text	text, markdown, html	Return plain text, basic Markdown, or escaped HTML. Line breaks are kept. The source formatting may be simplified.	"markdown"	Keep text.
descriptionMaxLength	Description length limit	integer	No	None	0 or more	Trim the returned description to this many characters. Leave blank for the full public text.	500	Use for smaller exports.
compact	Compact output	boolean	No	Off	true, false	Leave out the nested requirements and skill IDs. Other fields stay the same.	true	Use for smaller exports.
excludeEmptyFields	Hide empty fields	boolean	No	Off	true, false	Leave out fields whose value is null or an empty string.	true	Keep off for fixed CSV columns.
query	Single search phrase	string	One of the source fields	None	Any	A single search phrase. When set, it takes the place of Search phrases.	"python"	Use queries in the form.
rawUrl	Single search URL	string	One of the source fields	None	Public upwork.com search link	A public Upwork job search link. Its filters take the place of the same input filters.	"https://www.upwork.com/nx/search/jobs/?q=python"	Use startUrls for many links.
sort	Search order	string	No	Default: recency	recency, relevance+desc	Read newest jobs first, or use the source relevance order. Large searches may be split by category.	"recency"	Keep recency for alerts.
perPage	Jobs per search page	integer	No	Default: 50	1–50	Read this many jobs per page. A smaller page reads less data when you need only a few jobs.	20	Keep 50.
page	First search page	integer	No	Default: 1	1–101	Start from this page. Page 1 starts at the newest matching jobs.	2	Keep 1.
pagesToScrape	Search pages per input	integer	No	None	1 or more	Optional page limit for each search. Leave blank to keep paging until the result limit or the end.	3	Use to cap source requests.
subcategories	Subcategory IDs	array	No	None	Any	Upwork subcategory IDs from a search link.	ID from a link	Copy from a link.
skillIds	Skill IDs	array	No	None	Any	Upwork skill IDs from a search link.	ID from a link	Copy from a link. Use requiredSkills for names.
hourlyRateRange	Hourly range from search link	string	No	None	Any	Use an Upwork hourly range such as 20-50. For a simple minimum or maximum, use the hourly rate fields below.	"20-50"	Use minHourlyRate unless you copy a link.
fixedPriceRange	Fixed budget ranges from search link	string	No	None	Any	Use Upwork budget ranges such as 100-499,500-999. For a simple minimum or maximum, use the budget fields below.	"100-499,500-999"	Use minBudget unless you copy a link.
customFilters	Custom field filters	array	No	None	Operators: includes, notIncludes, equals, notEquals	Filter public fields with includes, notIncludes, equals or notEquals. Use objects with key, operator and value. Filters are combined with AND.	[{"key":"jobType","operator":"equals","value":"hourly"}]	Use for rare needs.
concurrency	Jobs at a time	integer	No	Default: 4	1–8	Read up to this many jobs together. The Actor waits when the source asks it to slow down.	2	Lower to 1 or 2 if you see rate limits.

categories uses Upwork IDs. In the form you pick the name. In the API use the IDs from the form or from an Upwork search link. customFilters[].key accepts title, description, jobType, experienceLevel, budget, tags, skills, absoluteDate, paymentVerified, clientLocation, clientCountry, allowedApplicantCountries, clientRating, clientTotalSpent and proposals.

Interacting settings

When	Input A	Input B	Effective behavior	Example	Cost/result effect
Both search fields are filled	query	queries	query takes the place of queries.	query: "python" and queries: ["java"] search only for python.	Fewer searches, fewer source requests.
A link has a filter and the same input is filled	startUrls or rawUrl	Same filter, such as jobType	The filter in the link wins for that link.	A link with t=1 keeps fixed jobs even with jobType: ["hourly"].	Rows follow the link.
A client filter is on and details are off	minClientRating, minClientTotalSpent, clientCountries, paymentVerified and similar	includeDetails: false	The run stops with DETAIL_FILTER_NEEDS_DETAILS.	includeDetails: false with paymentVerified: true.	One start event, no rows.
You need edits as well as new jobs	onlyNew: true	emitChanged: true	New jobs and changed jobs are saved. changeType tells which.	onlyNew: true, emitChanged: true.	Changed rows are charged as saved jobs.
You choose monitor mode	mode: "monitor"	onlyNew, emitChanged	Both turn on. You can set emitChanged: false.	mode: "monitor".	The first run saves all matches. Later runs save only new or changed jobs.
You run two separate monitors	stateName	onlyNew or skipReposts	Each name keeps its own history. Without a name, all runs share upwork-jobs-scraper-state.	stateName: "python-jobs".	A new name starts over, so the first run saves all matches.
You restart a history	resetState: true	onlyNew	The run ignores the old history. The new one has only this run's jobs.	resetState: true once, then remove it.	Saves all matches again, and they are charged again.
Both age filters are filled	maxJobAge	maxAgeMinutes	maxJobAge wins.	maxJobAge: {"value": 2, "unit": "days"}.	Older jobs are filtered and not charged as saved jobs.

Input examples

Smallest search

{
  "queries": ["web scraping"],
  "maxItems": 10
}

Expected behavior: reads the newest jobs for the phrase, with public details. Expected output: up to 10 job rows. Limit or cost note: about $0.0106 on Free, $0.0079 on Gold.

Well-paying clients only

{
  "queries": ["n8n automation"],
  "maxItems": 100,
  "jobType": ["hourly"],
  "minHourlyRate": 30,
  "paymentVerified": true,
  "minClientTotalSpent": 10000
}

Expected behavior: searches, reads details, then keeps only matching jobs. Expected output: up to 100 rows. A narrow filter may give fewer. Limit or cost note: rows that do not match are free, but the source requests used to check them are paid.

New job alerts on a schedule

{
  "queries": ["web scraping"],
  "maxItems": 50,
  "onlyNew": true,
  "stateName": "web-scraping-alerts",
  "deliveryConnector": "<your connector>",
  "deliveryTarget": "<Slack channel ID or Notion page link>"
}

Expected behavior: the first run saves up to 50 jobs as a starting point. Later runs save only jobs not saved before. Expected output: new jobs, plus a short summary in Slack or Notion. Limit or cost note: a run with no new jobs still pays for the start and source requests.

Known jobs by link or ID

{
  "mode": "details",
  "jobIds": ["2117261165716622937"],
  "maxItems": 1
}

Expected behavior: reads public details for this job. Expected output: one row. A job that is hidden from visitors gives no row and adds to membersOnly in RUN_SUMMARY. Limit or cost note: no row means no job event.

👉 Try it now: open the input form

Output

The Dataset has one row per saved job. The default table shows the main columns. All fields are in the JSON, CSV and API exports. RUN_SUMMARY in the default key-value store has the counts and stop reason.

Output fields

Field	Meaning	Type	Can be empty?	Example / enum	Source / notes
jobId	Upwork job ID.	string	No	"2117926027372106681"	Job post
id	The same ID as jobId, for imports.	string	No	"2117926027372106681"	Job post
ciphertext	Upwork ID used in the job link.	string	No	~022117261165716622937	Job post
title	Job title, with contact details removed.	string	No	See example below	Job post
description	Full public job text, with email addresses and phone-like numbers removed.	string	Yes	Full job text	Job post
budget	Fixed-price budget. Null for hourly jobs or an undisclosed budget.	number	Yes	10000	Job post
budgetCurrency	Currency code of the fixed budget.	string	Yes	"USD"	Job post
hourlyRange	Known minimum and maximum hourly rates. Null limits mean undisclosed.	object	Yes	{"min": 6, "max": 25}	Job post
hourlyBudgetMin	Known minimum hourly rate.	number	Yes	6	Job post
hourlyBudgetMax	Known maximum hourly rate.	number	Yes	25	Job post
weeklyRetainerBudget	Weekly retainer budget, when public.	number	Yes	Null	Job post
jobType	hourly, fixed or weekly_retainer.	string	Yes	hourly, fixed, weekly_retainer	Job post
experienceLevel	entry, intermediate or expert.	string	Yes	entry, intermediate, expert	Job post
skills	Public skill names.	array	Yes	["Data Extraction Tools", "Software Debugging"]	Job post
skillIds	Public skill IDs.	array	Yes	["1253342873018576896", "1031626781403627520"]	Job post
duration	Project duration label.	string	Yes	"Less than 1 month"	Job post
durationWeeks	Project duration in weeks.	number	Yes	3	Job post
workload	Weekly work requirement.	string	Yes	"Less than 30 hrs/week"	Job post
postedAt	Publication time reported by Upwork.	string	Yes	"2026-11-04T10:24:51.983Z"	Job post
absoluteDate	The same publication time as postedAt.	string	Yes	"2026-11-04T10:24:51.983Z"	Job post
createdAt	Time the job was created.	string	Yes	"2026-11-04T10:24:49.148Z"	Job post
renewedAt	Source renewal or sourcing time, if given.	string	Yes	"2026-11-04T09:55:21.845Z"	Job post
clientCountry	Client country name. Source country codes are converted to country names. Null when hidden.	string	Yes	"United States"	Public job details
clientCountryCode	Two-letter ISO country code from the public client country name. Null when the name cannot be matched.	string	Yes	"US"	Actor
clientCity	Public client city, when visible.	string	Yes	City name	Public job details
clientTimezone	Public client timezone label.	string	Yes	"America/Chicago (UTC-06:00)"	Public job details
clientPaymentVerified	Whether the public page shows a verified payment method.	boolean	Yes	false	Public job details
paymentVerified	The same value as clientPaymentVerified.	boolean	Yes	false	Public job details
clientTotalSpent	Public lifetime client spend. A hidden amount stays null.	number	Yes	17584.47	Public job details
clientSpentCurrency	Currency of the public client spend.	string	Yes	"USD"	Public job details
clientRating	Public client rating. Null when the client has no reviews.	number	Yes	5	Public job details
clientReviewCount	Public review count.	number	Yes	0	Public job details
clientTotalHires	Public total client assignments.	number	Yes	10	Public job details
clientActiveHires	Public active assignments.	number	Yes	3	Public job details
clientTotalHours	Public hours billed.	number	Yes	0	Public job details
clientMemberSince	Public client account creation date.	string	Yes	"2021-03-17T00:00:00.000Z"	Public job details
clientIndustry	Public client industry. Null when empty.	string	Yes	"Health & Fitness"	Public job details
clientCompanySize	Company size group reported by Upwork. The group code is not a staff count.	number	Yes	10	Public job details
clientHireRatePercent	Null when this value is hidden from visitors.	number	Yes	Null	Public job details
clientAvgHourlyRate	Null when this value is hidden from visitors.	number	Yes	Null	Public job details
category	Broad job category.	string	Yes	"Data Science & Analytics"	Job post
subcategory	Job subcategory.	string	Yes	"Data Extraction/ETL"	Job post
contractToHire	Whether the job is marked contract-to-hire.	boolean	Yes	false	Job post
totalApplicants	Exact public applicant count. Null means unknown.	number	Yes	18	Public job details
proposals	The same applicant count as totalApplicants.	number	Yes	18	Public job details
interviewing	Public count of invited interviews.	number	Yes	3	Public job details
invitesSent	Public invitations sent.	number	Yes	5	Public job details
unansweredInvites	Public unanswered invitations.	number	Yes	2	Public job details
hired	Public hires on this job.	number	Yes	0	Public job details
positionsToHire	Number of positions on this job.	number	Yes	0	Public job details
clientLastActivity	Last client activity time shown on the job page.	string	Yes	"2026-11-04T10:27:50.616Z"	Public job details
allowedApplicantCountries	Applicant country preferences from the public job requirements.	array	Yes	["United States"]	Public job details
qualifications	Public freelancer requirements. Missing values stay null; ANY or 0 means no requirement.	object	Yes	See note below	Public job details
jobStatus	open or closed, from the public page.	string	Yes	open, closed	Public job details
detailsStatus	available, not_requested, members_only, not_found or unavailable (the source gave no usable answer twice, so the row keeps search fields only).	string	No	available, not_requested, members_only, not_found, unavailable	Actor
source	Source website.	string	No	"https://www.upwork.com"	Actor
url	Public Upwork job link.	string	No	"https://www.upwork.com/jobs/~022117926027372106681"	Job post
query	Search phrase that found the job. Null for a direct lookup.	string	Yes	"web scraping"	Actor
fetchedAt	UTC time when the job was read.	string	No	"2026-11-04T12:08:00.361Z"	Actor
firstSeenAt	First saved time in this monitor history, or this run.	string	Yes	"2026-11-04T12:08:00.361Z"	Actor
lastSeenAt	Last time this row was saved.	string	Yes	"2026-11-04T12:08:00.361Z"	Actor
changeType	new or changed, based on the monitor history.	string	Yes	new	Actor
tags	The same public skill names as skills.	array	Yes	["Data Extraction Tools", "Software Debugging"]	Job post
clientLocation	The same public country as clientCountry.	string	Yes	"United States"	Public job details
clientCountryRaw	Country text returned by Upwork before conversion.	string	Yes	"United States"	Public job details

qualifications has countries, regions, englishLevel, freelancerType, minHours, minHoursPerWeek, minJobSuccess, portfolioRequired and risingTalent. ANY or 0 means no requirement.

Output entities and surfaces

Situation	Dataset/API/KVS/live view	Result shape	User action
Normal run	Dataset rows, RUN_SUMMARY status: "complete"	One row per saved job	Use the rows.
No jobs match, or no new jobs	Empty Dataset, RUN_SUMMARY status: "not_found" with filtered or unchanged counts; run SUCCEEDED	No rows	Change the phrase or filters. For a monitor this is normal.
Job hidden from visitors	No row for a direct lookup, status: "restricted", membersOnly count. Search rows keep search fields with detailsStatus: "members_only"	Fewer fields	Open the url while logged in on Upwork.
Your cost limit is reached	Saved rows stay, partial: true, stopReason: "CHARGE_LIMIT"; run SUCCEEDED	Partial rows	Raise the limit and run again.
Upwork gives no usable details for one job	The row keeps search fields with detailsStatus: "unavailable"; a direct lookup lists the job in unprocessed	Fewer fields	Run that job again later.
Source block, too long a wait, timeout or stop	Saved rows stay, partial: true, stopReason tells why; run FAILED	Partial rows or none	See Troubleshooting.
Bad input	No rows, status: "error", stopReason tells why; run FAILED	No rows	Fix the input.

Output example

Real row, some fields left out:

{
  "jobId": "2117702916010329632",
  "title": "Social Media automated data collection (API) and Website Data Collection",
  "jobType": "hourly",
  "hourlyRange": {
    "min": 6,
    "max": 25
  },
  "budget": null,
  "experienceLevel": "intermediate",
  "skills": [
    "Data Scraping",
    "Web Scraping",
    "API Integration",
    "Social Media Marketing"
  ],
  "duration": "1 to 3 months",
  "postedAt": "2026-11-03T19:38:23.701Z",
  "clientCountry": "United States",
  "clientPaymentVerified": true,
  "clientTotalSpent": 599.1,
  "clientSpentCurrency": "USD",
  "clientRating": 5,
  "clientReviewCount": 2,
  "clientTotalHires": 5,
  "totalApplicants": 37,
  "jobStatus": "open",
  "detailsStatus": "available",
  "url": "https://www.upwork.com/jobs/~022117702916010329632",
  "fetchedAt": "2026-11-04T12:08:00.362Z"
}

Send results to Slack or Notion

Pick your connector in the optional deliveryConnector field. After the run saves its data, the Actor posts a short summary (up to 25 rows and a link to the full dataset) there. Put a Notion page link or a Slack channel ID in deliveryTarget: in Notion every run is added on top of that one page, so your workspace does not fill up with new pages. A delivery error never fails the run and costs nothing extra. Connect a service once in Apify Console → MCP connectors.

Use case → fields

Use case	Fields to read first
Lead list	title, clientCountry, clientTotalSpent, clientRating, clientPaymentVerified, url
Job alert	title, budget, hourlyRange, skills, postedAt, changeType, url
Market research	category, subcategory, skills, experienceLevel, hourlyBudgetMin, hourlyBudgetMax, budget
Competition check	totalApplicants, interviewing, invitesSent, hired, clientLastActivity

👉 Try it now: open the input form

Pricing

Pricing unit

You pay per event. Apify platform usage is included in these prices.

Event	When	Free	Bronze	Silver	Gold and up
Run start (apify-actor-start)	Once per GB of run memory, at least once. The default 256 MB is one event.	$0.001	$0.001	$0.001	$0.001
Source request (source-page)	One search page or one batch of up to 6 job details	$0.0001	$0.0001	$0.0001	$0.0001
Saved job (job)	One job row saved to the Dataset	$0.00090	$0.00081	$0.00072	$0.00063

• No results: you pay for the start and source requests.
• A saved row with fewer fields (detailsStatus is members_only or unavailable) is charged as a saved job.
• Filtered jobs and unchanged monitor jobs are not charged as saved jobs. The source requests used to read them are paid.
• A partial run pays only for the work it did. A failed request never costs more than a successful one.
• Use Maximum cost per run (maxTotalChargeUsd in the API) to set a limit. The smallest limit is $0.0022. When the limit is reached, the run keeps its saved rows and ends as SUCCEEDED with stopReason: "CHARGE_LIMIT".

Cost examples

Measured source requests. Your count depends on the search and on how many jobs have details.

Scenario	Input scope	Expected item/event count	Base charge	Optional charge	Stop/cap behavior
First test	1 phrase, maxItems: 10	1 start, 6 source requests, 10 jobs	$0.0106 Free, $0.0079 Gold	None	Stops at 10 jobs
Small list	1 phrase, maxItems: 100	1 start, about 29 source requests, 100 jobs	$0.0939 Free, $0.0669 Gold	None	Stops at 100 jobs
Market sample	1 phrase, maxItems: 1000	1 start, about 284 source requests, 1,000 jobs	$0.9294 Free, $0.6594 Gold	None	Stops at 1,000 jobs or your cost limit
Large export	Broad phrase, maxItems: 6000	1 start, about 1,215 source requests, 6,000 jobs	$5.5225 Free, $3.9025 Gold	None	Needs a longer timeout
Monitor with no new jobs	onlyNew: true, 1 job checked	1 start, 1 source request, 0 jobs	$0.0011	None	Ends with no rows

👉 Try it now: open the input form

Limits

• Only public data. There is no login, so private and invite-only jobs are not available.
• Some jobs show details only to members. In our 6,000-job test, 1,055 jobs had no public details. These rows keep the search fields and have detailsStatus: "members_only".
• clientAvgHourlyRate and clientHireRatePercent are hidden from visitors, so they stay null.
• The largest tested run saved 6,000 jobs in about 20 minutes. The default timeout is 300 seconds. Raise it for runs over about 1,000 jobs.
• A big search is split by category, type and experience level to get past the source limit of about 5,000 results per search. So a big result has no single global order.
• If the source still holds back part of a search, the summary says SOURCE_CAP and sourceCapPartitions counts the affected parts.
• skipReposts skips only exact repeats of the cleaned title and description. Reworded reposts are saved.
• A monitor history keeps up to 100,000 jobs. Closed and then the oldest jobs are dropped first.
• title and description have email-like and phone-like text replaced with [contact removed].
• You are responsible for how you use the data and for following Upwork's Terms and the laws that apply to you.

👉 Try it now: open the input form

API, export and integrations

• API: start a run with POST https://api.apify.com/v2/acts/cliqtomedia~upwork-jobs-scraper/runs, using the same JSON as the form. Read rows from https://api.apify.com/v2/datasets/<defaultDatasetId>/items?view=jobs. Leave out view=jobs to get all fields. Read the report from RUN_SUMMARY in the default key-value store.
• Export: download the Dataset as JSON, CSV, Excel or XML from the Output tab or the API.
• Scheduling: create an Apify Schedule with onlyNew: true. The next system reads jobId, changeType and url.
• Webhooks / no-code: use an Apify webhook on run success, then read the Dataset with Make, Zapier or n8n. Use jobId as the unique key.
• Slack / Notion: set deliveryConnector, see Send results to Slack or Notion.

Use with AI agents (MCP)

• MCP server URL with only this Actor: https://mcp.apify.com?tools=cliqtomedia/upwork-jobs-scraper.
• Smallest input: {"queries": ["web scraping"], "maxItems": 10}.
• Key fields: title (job name), hourlyRange or budget (pay), clientTotalSpent (public lifetime client spend), totalApplicants (exact public count), url (job link).
• Typical run: 10 jobs in about 6 seconds, up to about a minute if Upwork refuses the first requests, about $0.0106 on the Free plan.
• Limits: public data only, keep maxItems small in agent chats, and read stopReason in RUN_SUMMARY when the Dataset has fewer rows than asked.

👉 Try it now: open the input form

Troubleshooting

The run has no rows

Symptom: the run is SUCCEEDED and the Dataset is empty. Likely reason: no job matched, or a monitor found nothing new. Check: scanned, filtered and unchanged in RUN_SUMMARY. Action: use fewer filters or a wider phrase. For a monitor, wait for new jobs.

The search returned jobs that do not match my phrase

Symptom: a very unusual phrase returns a few jobs that look unrelated. Likely reason: Upwork's own search also returns loosely related posts. Check: title and description in the rows. Action: add includeKeywords or titleContains to keep only jobs with your exact words.

The run stops with DETAIL_FILTER_NEEDS_DETAILS

Symptom: the run is FAILED and has no rows. Likely reason: a client or applicant filter is on, but includeDetails is false. Check: stopReason in RUN_SUMMARY. Action: set includeDetails to true, or remove the filter.

The run saved fewer jobs than maxItems

Symptom: fewer rows than maxItems. Likely reason: the search ended, filters removed jobs, or the run stopped early. Check: saved, filtered, partial and stopReason in RUN_SUMMARY. Action: for CHARGE_LIMIT, raise the cost limit. For DEADLINE, raise the timeout. For SOURCE_CAP, narrow the search by category or date.

The run failed with RATE_LIMIT_WINDOW or SOURCE_BLOCKED

Symptom: the run is FAILED. Rows saved before the stop stay in the Dataset. Likely reason: Upwork asked for a long wait or refused the request with a browser check. The Actor waits, opens a new connection and tries up to twice more. It does not solve these checks. How often: in our short test runs Upwork refused the first requests in 25 of 68 runs. The Actor recovered in 23 of those 25, and these runs took about 15–50 seconds instead of about 6. The other 2 runs (about 3% of all) ended with SOURCE_BLOCKED and cost about $0.0013. Starting the run again usually works. In our latest runs of 1,000 and 6,000 jobs Upwork refused nothing. Check: stopReason, rateLimits, retries and sourceRefusals in RUN_SUMMARY. Action: run again later, and lower concurrency to 1 or 2. This does not mean that there are no jobs.

Fields are empty for some jobs

Symptom: client fields are null and detailsStatus is members_only, not_requested or unavailable. Likely reason: Upwork shows these details only to members, includeDetails is off, or Upwork gave no usable answer for this job twice. Check: detailsStatus in the row. Action: keep includeDetails on. For a members_only job, open url while logged in on Upwork. A monitor remembers a job with empty fields: it sends the job again with full details only when emitChanged is on. To get an unavailable job right away, run it by its link or ID.

The history name is not available

Symptom: warnings in RUN_SUMMARY has STATE_NAME_UNAVAILABLE. Likely reason: this name is used by a store that this run cannot open. Check: stateStoreId in RUN_SUMMARY. Action: nothing is needed, as the Actor uses a similar name. To control the name, set a new stateName.

👉 Try it now: open the input form

FAQ

Do I need an Upwork account or cookies?

No. The Actor reads only what Upwork shows to visitors. Do not give it your login.

Can I scrape Upwork jobs by a search link?

Yes. Run your search on Upwork, copy the link and paste it in startUrls. Filters in the link are used.

How do I get only new jobs?

Set onlyNew: true and run the Actor on an Apify Schedule. The first run saves all matches up to maxItems as a starting point. Later runs save only jobs not saved before.

Why do I pay for source requests when no job is saved?

The Actor must read the source to find out if a job matches. So each source request is charged, even if your filters remove all jobs. Make filters part of the search when you can, for example jobType or experienceLevel.

Why is clientRating null?

The client has no public reviews yet. null means that the value is unknown or hidden, and it is not the same as 0.

How fresh is the data?

Each row is read during the run, and fetchedAt shows the time. Rows are not updated later. Run again or use a schedule for fresh data.

Can this Actor apply for jobs or contact clients?

No. It only reads public job posts.

Is this an official Upwork product?

No. This Actor is not affiliated with or endorsed by Upwork.

👉 Try it now: open the input form

Related Actors

• Remote Jobs Aggregator - RemoteOK, WWR & Himalayas, No login: add remote jobs from other public boards.
• Wellfound Jobs Scraper: Startup Jobs to CSV & JSON: get startup jobs when you look for full-time roles.
• Google Jobs Scraper - Salary, Apply Links & Dates, No Login: search jobs from many job sites.

👉 Try it now: open the input form

Support

Open the Issues tab of this Actor. Include the run ID, your input without private data, and the stopReason from RUN_SUMMARY.

This Actor is not affiliated with or endorsed by Upwork.

👉 Try it now: open the input form