Overview
JobsWithPay is a technology job aggregator built around a single organizing principle: compensation should be visible before a candidate spends an hour on an application. The platform currently indexes roughly 93,000 open technical roles sourced from the career pages and applicant tracking systems of 7,793 companies, and it reports that about 44.3 percent of those listings state a pay range. Rather than reselling recruiter submissions or scraping third party mirrors, the service reads each posting from the employer's own board, whether that board runs on Greenhouse, Ashby, Lever, Workday, or a bespoke careers site.
The audience is narrow by design. Software engineers, machine learning and inference specialists, data scientists, infrastructure and security practitioners, quant researchers, product managers, and designers make up the bulk of the taxonomy. The implied user is a candidate who wants to filter by money first and title second, and who is tired of discovering a salary band only after three interview rounds.
Market positioning is openly adversarial toward LinkedIn. A published side by side comparison argues that LinkedIn discloses pay on 18.7 percent of the AI and ML results sampled, that 26.7 percent of those results were not actually AI or ML jobs, that 3.3 percent came from staffing agencies, and that 8.4 percent were duplicate copies of the same role. JobsWithPay counters with a deduplicated index, a strict company-only sourcing rule, and a daily recheck cycle. Competing references in this space include Levels.fyi for compensation benchmarking, Wellfound and Otta for startup discovery, and Hired for matching driven search. JobsWithPay's differentiator is the combination of a live job index, salary charting, and a portfolio layer that tries to prove skills rather than list them.
Key Features
Company-sourced listings with agency noise excluded. Every posting originates from the hiring company's own board or careers site. The stated policy bars staffing agencies and sites that copy jobs, and each role appears exactly once. For a handful of large employers whose careers sites cannot be read directly, the company purchases a job data feed, which is disclosed openly on the freshness panel rather than hidden.
Posted pay, never estimated. Salary figures are the numbers the employer published, reproduced without adjustment or modeling. Where no pay is posted, the listing simply reads "Pay not posted" instead of guessing a band. Charts use annual base pay ranges in US dollars, and the interface repeatedly notes that base pay excludes stock and bonus, a distinction that matters at equity heavy employers.
Pay explorer with company ranking. A charting tool lets a visitor choose a job type (software, ML and AI, data science, product, design) and a place (Bay Area, New York, Remote in the US), then ranks employers by the median low end to median high end of their posted ranges. Netflix leads the Bay Area software view at a $388K to $558K posted range across 23 jobs, followed by Anthropic and Roku. A company needs at least five jobs with posted pay to qualify, which filters out statistical noise. The full dataset lives in the pay explorer.
AI search in plain language. The search bar accepts a sentence rather than keywords, such as remote senior ML roles paying $200K or more. Example prompts on the homepage include new graduate software jobs in New York with visa sponsorship and data engineer jobs in Bengaluru that use Spark. The system reads the description, tags each job before display, and returns matches, which is the mechanism behind the claim that zero percent of results are mistagged.
Daily re-verification with visible timestamps. Each job is re-read from the company board once a day, and a role missing from two consecutive checks is removed. The homepage reports that 86,575 of 93,089 open jobs, about 93 percent, were verified within the previous 24 hours. A source table breaks coverage down by ATS, showing 36,187 roles from Greenhouse, 21,875 from Ashby, 10,685 from company careers sites, 9,507 from Lever, 6,146 from Workday, plus smaller feeds from Y Combinator, Hacker News hiring threads, and Workable. A coverage page lists the underlying employers.
Portfolio pages that map proof to requirements. Users can generate a public page from a resume, and the platform enriches it with evidence pulled from public work such as merged pull requests, published papers, packages, and repositories. Each skill is labeled Proven, Stated, or No proof. A recruiter view overlays a specific job's requirements against that evidence, so an inference engineer role might show three proven skills, one stated skill, and one gap, with the supporting artifact named in each case.
Structured job taxonomy and regional hubs. Fifteen job families, from software engineering and inference systems to quant trading, hardware and robotics, and solutions engineering, each carry dedicated landing pages for the Bay Area, New York, remote, India, Bengaluru, Hyderabad, Delhi NCR, Pune, and Mumbai. This gives the site unusually deep Indian market coverage for a US oriented product.
How It Works
A visitor can search immediately without an account. The homepage presents a single sentence style input, a row of quick filters for remote, posted pay, entry level, AI and ML, and posted today, and a scrolling list of the newest roles that include compensation, each showing the company, the pay band, and the date of the last check.
Selecting a listing opens a detail page with the posted range, the source board, and the verification timestamp, and the apply action hands off to the employer's own site rather than an internal application form. This keeps the candidate inside the company's real funnel and removes the resume parsing step that aggregators often add.
Signing in requires only an email address. That unlocks saved jobs, email alerts, the AI tools, and portfolio creation. The portfolio flow starts from a resume upload, then asks for public work links. The system scans those sources, scores each skill claim, and produces a shareable page plus a printable resume derived from it.
Daily operation is passive. There is no application tracker, no inbox, and no recruiter messaging layer, which keeps the product closer to a search engine than to a job CRM. The trade-off is that follow up and pipeline management must happen elsewhere.
Use Cases
A senior ML engineer benchmarking a market. A candidate weighing two offers in the Bay Area can open the pay explorer, select ML and AI plus Bay Area, and see which employers post the widest ranges. Comparing a $320K to $485K median band at Anthropic against a $266K to $445K band at OpenAI gives a concrete negotiating anchor rather than a rumor.
A new graduate filtering by sponsorship and pay. A student searching for entry level software roles with visa sponsorship can phrase the query in natural language, then apply the posted pay filter to skip listings that hide compensation. The entry level and new graduate taxonomies narrow the set further.
An infrastructure hire relocating to India. Because the taxonomy includes Bengaluru, Hyderabad, Delhi NCR, Pune, and Mumbai for every job family, a platform engineer can compare local postings against remote US roles without rebuilding the search from scratch.
A bootcamp graduate building credible evidence. Someone without a conventional pedigree can generate a portfolio page where public repositories and released packages count as proof. The Proven, Stated, No proof labels give a recruiter a fast read and give the candidate a specific list of gaps to close.
A recruiter doing a one click fit check. The sample portfolio page demonstrates a hiring manager pasting a job description and seeing which requirements are backed by artifacts. That reduces screening time for technical roles where resume claims are hard to verify.
Pricing & Value
Every feature is currently free while the product is in beta. That includes the full job index, the AI search tools, the pay explorer, job email alerts, interview question pages, and portfolio pages. Authentication is email only, with no credit card step disclosed anywhere in the flow.
No paid tier, seat pricing, or enterprise plan is published, and there is no indication of what will be charged once beta ends. Against alternatives, the value proposition is straightforward. Levels.fyi offers deeper historical compensation data and verified submissions but is not a live job board. LinkedIn offers enormous volume but, by JobsWithPay's own measurement, discloses pay far less often and mixes in agency listings. For a free beta product, the coverage and freshness numbers are strong. The main risk is durability: a free product with no visible revenue model may change terms, throttle features, or monetize employer listings later.
Final Verdict
JobsWithPay is a focused, technically credible entrant in a crowded category. Its strongest asset is the sourcing discipline: reading only company boards, deduplicating, and rechecking daily produces an index that feels cleaner than most aggregators, and the visible timestamps make that claim auditable rather than aspirational. The pay explorer is genuinely useful for negotiation research, and the portfolio concept addresses a real weakness in resume screening.
The limitations are equally clear. Base pay only, no equity or bonus modeling, means the numbers understate total compensation at exactly the companies that pay best in stock. Coverage leans on a specific set of ATS platforms and US dollar ranges. And the entire product sits behind a beta label with no pricing, which is a caution flag for anyone building a workflow on top of it.
This suits engineers, researchers, and data practitioners who prioritize disclosed compensation and want a low noise search experience. Candidates who need equity modeling or deep compensation history should pair it with Levels.fyi. Recruiters wanting outreach tooling should look at a sourcing platform instead. For pure pay transparent job discovery, JobsWithPay is worth a bookmark.
