The Upwork Job Market in July 2026: 161,102 Postings, Analyzed
Vibeworker observed 161,102 deduplicated public Upwork job postings in July 2026 — up 26% from June's 127,607. That is roughly 5,200 new jobs a day, or one every 16.6 seconds, around the clock.
This report describes what those postings looked like: how they paid, what experience level clients asked for, which AI tools they named, and which categories carried the volume. It measures postings, not outcomes—not proposals sent, contracts awarded, or final negotiated pay.
The headline is the pay, and it got worse. Fixed-price budgets skewed lower than June across the board, and the effective hourly rate implied by those budgets fell with them.
Fixed-price budgets skewed sharply lower
Of July's postings, 77,696 were fixed-price with a listed budget, 55,162 were hourly, and the remainder did not specify a type. Among the fixed-price jobs, the median budget was $80, down from $100 in June.
| Budget band | Jobs | Share |
|---|---|---|
| Under $50 | 32,550 | 41.9% |
| $50–$100 | 6,709 | 8.6% |
| $100–$250 | 14,118 | 18.2% |
| $250–$500 | 10,790 | 13.9% |
| $500–$1,000 | 6,032 | 7.8% |
| $1,000+ | 7,497 | 9.6% |
Nearly 42% of fixed-price postings budgeted under $50, up from 31.5% in June. Jobs at $500 or more—the band where a project can pay for meaningful time—made up just 17.4% of the fixed-price feed, down from 22.3%.
Dividing each fixed-price budget by Vibeworker's LLM estimate of the hours the job would take gives an effective hourly rate. The median fixed-price job in July worked out to about $6.25 an hour, down from $8.75 in June. That is a listed-budget figure, not completed-contract pay, and budgets are often negotiable—but it is what clients placed in front of applicants, and the direction is the same as the raw budget numbers: down.
Clients asked for experience they were not budgeting for
The experience tier clients selected did not match the pay they attached, and the mix barely moved from June. Across July postings with an experience level set:
| Experience level | Jobs | Share |
|---|---|---|
| Intermediate | 85,068 | 67.3% |
| Expert | 30,890 | 24.4% |
| Entry level | 10,491 | 8.3% |
Only about one posting in twelve was labeled Entry level—effectively unchanged from June's 8.3%. Roughly two-thirds asked for Intermediate talent, and a quarter asked for Experts, while the median fixed budget fell to $80. The market openly asks for experience far more often than it pays for a beginner to gain it, and July's lower budgets make that gap more pronounced, not less.
ChatGPT led AI-tool demand growth as most terms cooled
Vibeworker tracks how often specific AI and automation tools appear in job titles, normalized per 100,000 postings so that month-to-month volume changes do not distort the comparison. In July, most tracked terms continued June's decline, with ChatGPT and Gemini the exceptions.
| Tool | June per 100k | July per 100k | Change |
|---|---|---|---|
| ChatGPT | 58.8 | 73.9 | +25.7% |
| Gemini | 34.5 | 38.5 | +11.6% |
| Replit | 35.3 | 33.5 | −5.1% |
| AI agent | 269.6 | 255.1 | −5.4% |
| OpenAI | 88.6 | 82.6 | −6.8% |
| n8n | 283.7 | 262.6 | −7.4% |
| Claude | 433.4 | 383.0 | −11.6% |
| Zapier | 133.2 | 86.9 | −34.8% |
ChatGPT reversed its June slide, up 25.7% in title mentions, and Gemini kept climbing. Everything else eased further, led down by Zapier, which lost more than a third of its June rate. Two results held steady: Claude remained the single most-named AI tool in Upwork titles by a wide margin (617 raw mentions, 383 per 100,000) despite a further decline, and n8n stayed the most-named automation platform ahead of Zapier and Make.com.
The chart and table apply a minimum threshold of 40 raw title mentions in at least one month, so a term moving from three mentions to six is not shown as triple-digit growth.
Websites, IT & Software led July's job supply
| Category | Jobs | Share |
|---|---|---|
| Websites, IT & Software | 23,487 | 14.6% |
| Digital Marketing | 16,398 | 10.2% |
| Video & Animation | 14,998 | 9.3% |
| Graphic, Editorial & Presentation Design | 10,889 | 6.8% |
| Web Development | 8,232 | 5.1% |
| Lead Generation & Telemarketing | 8,211 | 5.1% |
| Web & Mobile Design | 6,610 | 4.1% |
Marketing and creative categories still carried real volume—Digital Marketing and Video & Animation together accounted for close to a fifth of postings—but the largest single category in July's feed was software and web work.
One more signal: unverified-payment postings jumped
Payment verification is one of the cheapest quality checks a freelancer can make before spending connects. In July, 39.6% of postings came from clients whose payment method was not verified, sharply up from 24.0% in June. Separately, 7.3% of scored postings tripped Vibeworker's red-flag screen (a score of 3 or below out of 10), up from 4.8%—both signals moving the same direction as July's lower budgets.
What the month says
July's feed grew 26% over June, and the extra volume skewed toward lower pay and weaker client signals: smaller fixed budgets, a lower effective hourly rate, more unverified clients, and a higher red-flag rate. Clients kept asking for experience they weren't budgeting for. For a freelancer, the practical consequence is the same as always, just sharper: the jobs worth a proposal are a shrinking minority of a growing feed, and filtering fast matters more than it did in June.
Methodology
Vibeworker analyzed 161,102 deduplicated public Upwork job postings observed from July 1 through July 31, 2026 UTC. Deduplication keeps one record per job.
Budget statistics use postings with a usable positive displayed budget; records without one were excluded from pay calculations. The effective hourly rate divides a posting's listed fixed budget by an LLM estimate of the hours the job would require, and is a listed-value figure rather than completed-contract pay. Experience-level shares use postings with a level set. AI-tool rates use literal case-insensitive title matches, are normalized per 100,000 postings, and are not mutually exclusive. Red-flag rates use Vibeworker's intrinsic per-job scoring.
This dataset covers postings observed by Vibeworker and should not be presented as an official Upwork census. It does not measure proposals, contracts awarded, final negotiated pay, or freelancer earnings.
Reuse and citation
You may quote these findings and republish the charts for editorial use. Please credit Vibeworker and link to this article as the original source. Charts may be resized but should not otherwise be modified.
Journalists working on a story can request a custom data cut — a specific category, date range, or metric not covered here — at no cost. Email mike@tryvibeworker.com with your outlet and press credentials.
Suggested citation: "Vibeworker, The Upwork Job Market in July 2026, August 2026."
Download the full dataset (JSON) →
Vibeworker tracks new Upwork jobs and sends you a notification the second one matches what you're looking for. Get started free →

Michael Watkins
Founder of Vibeworker. Helping freelancers win the Upwork game through speed and data.
Stop missing the jobs that matter
Vibeworker watches the Upwork feed and alerts you the moment a high-fit job appears — before the proposals pile up.
Get started free →Keep reading
The Upwork Job Market in June 2026: 127,607 Postings, Analyzed
A data report on 127,607 public Upwork job postings observed in June 2026: budget distribution, effective hourly pay, experience-level demand, AI-tool mentions, and category mix.
What Entry-Level Upwork Jobs Actually Pay
We analyzed 10,649 entry-level Upwork jobs. Three-quarters of fixed-price listings offered $100 or less, and 87% of visible hourly floors were $15 or less.
Upwork's Expert-Level Delusion, Visualized
We analyzed 28,559 expert-level Upwork jobs with visible pay. 44% of fixed-price listings offered $100 or less, while 53% of hourly floors opened at $15 or less.