Job postings API benchmark 2026: TheirStack, Bright Data and Crustdata
On job postings, TheirStack returned 76.6% of the open roles that companies listed on their own job boards, ahead of Bright Data at 63.1% and Crustdata at 50.4%.
On job postings, TheirStack returned 76.6% of the open roles that companies listed on their own job boards, ahead of Bright Data at 63.1% and Crustdata at 50.4%.
Three providers took the test. Each was asked about the same 458 open roles at 50 companies, and each answer was graded against the company's own job board. TheirStack returned 351 of 458 roles, Bright Data 289 and Crustdata 231.
A job postings API takes a company and returns the jobs it is hiring for, with titles, locations and dates. This test counts a hit only when the API returns the same job the company lists, with the same title, seniority and specialization, and the same city or country, or remote on both. A posting for a similar role at the same company counts as a miss.
The short version:
- TheirStack returned 76.6% of open roles, 13.5 points ahead of Bright Data at 63.1%.
- Crustdata is third at 50.4%, 12.7 points behind Bright Data. That gap is too small to call a clear difference.
- TheirStack led every age group, between 73.8% and 82.7% of roles.
Which API returns the most open roles?
TheirStack, with 76.6%. It returned 351 of the 458 open roles. Bright Data returned 289 and Crustdata 231.
| Provider | Roles returned | Roles (of 458) | Roles missed | Companies with no role found (of 50) |
|---|---|---|---|---|
| TheirStack | 76.6% | 351 | 107 | 2 |
| Bright Data | 63.1% | 289 | 169 | 9 |
| Crustdata | 50.4% | 231 | 227 | 14 |
TheirStack is 13.5 points ahead of Bright Data. Bright Data is 12.7 points ahead of Crustdata, a gap too small to call a clear difference. TheirStack is 26.2 points ahead of Crustdata.
Does the order change with the age of the role?
The order does not change with the age of the role. TheirStack returned the most roles in every age group, then Bright Data, then Crustdata. The table splits the 458 roles by how many days had passed since the company posted each one.
| Age of role | Roles | TheirStack | Bright Data | Crustdata |
|---|---|---|---|---|
| 0 to 7 days | 139 | 115 (82.7%) | 86 (61.9%) | 60 (43.2%) |
| 8 to 14 days | 145 | 107 (73.8%) | 97 (66.9%) | 77 (53.1%) |
| 15 to 30 days | 174 | 129 (74.1%) | 106 (60.9%) | 94 (54.0%) |
| All roles | 458 | 351 (76.6%) | 289 (63.1%) | 231 (50.4%) |
TheirStack returned the most roles in every age group. Bright Data stayed between 60.9% and 66.9%. Crustdata returned 43.2% of roles from the past 7 days, 53.1% of roles 8 to 14 days old and 54.0% of roles 15 to 30 days old. TheirStack's lead over Crustdata is 39.6 points on roles from the past 7 days and 20.1 points on roles 15 to 30 days old.
How was this benchmark measured?
Every provider was asked about the same open roles, and every answer was graded against the company's own job board.
- Panels. 458 open roles at 50 companies. The companies were drawn with a fixed random seed from published company lists, 29 from a startup directory and 21 from large-company rankings, and each one's job board was found from the careers page on its own website. Each company contributed up to 15 roles posted in the 30 days before the test.
- Input. Each provider got the same roles in the form its API takes.
- TheirStack and Crustdata received the company's website domain and the role's title, and returned up to 3 postings with that title from the past 90 days, by each API's own date field. TheirStack's date is the posting date, and its request kept only postings it marks as open. Crustdata's date is the day the posting was added, and its job search has no open-only filter. A lookup that returned nothing was sent once more without the seniority words.
- Bright Data's job scraper takes a company name, keywords and a location. It received the company's name, the same title words and the role's location, cleaned by a fixed rule that removes office words and internal codes and sets remote roles to their country. It returned up to 3 postings from any date. The API rejected 7 locations that still carried a two-letter country prefix, and those 7 were sent again without it. Other locations with a prefix were accepted as sent.
- What counts as a hit. A returned posting is the same job as the board role. That holds when the posting carries the board's own job id, or when model graders, which could not see which API answered, judged it the same job under one written standard. For Bright Data, whose postings name their own company, the posting also had to be at the same company. A second, independent grader re-judged 60 cases from TheirStack and Crustdata, and 40 from Bright Data, and agreed on all 100. A similar role with a different seniority, specialization or location counts as a miss.
- Score. Roles returned divided by all 458. Each job board was read before the first queries and again after the last queries, and a role counts only if it was listed at every reading. 7 of the 465 roles drawn had closed and were dropped for every provider.
- Providers. TheirStack, Bright Data and Crustdata.
The test does not measure price, speed, salary data, job descriptions or how fast a closed role disappears. 45 of the 50 companies are in the US, and every company publishes a public job board, so a list of companies in other regions, or of companies hiring only through job sites, can give a different order.
How do these providers compare on price and API access?
TheirStack sells monthly API credit plans that start at $49 for 1,500 credits, with 1 credit per job returned and a free plan of 200 API credits a month. Bright Data's job scraper is $1.50 per 1,000 records, pay as you go. Crustdata charges per credit, from $0.30 per credit, decreasing to $0.10 and lower on enterprise contracts.
| Basis | Bright Data | Crustdata | TheirStack |
|---|---|---|---|
| Open roles returned (of 458) | 63.1% | 50.4% | 76.6% |
| APIbenchmarks Index | 80.3 (B) | 95.7 (A+) | Not in the index |
| Pricing model | $1.50 per 1,000 records for the job scraper, pay as you go | From $0.30 per credit, decreasing to $0.10 and lower on enterprise contracts | Monthly API credit plans, 1 credit per job returned |
| Free access | 5,000 free credits a month | Free trial credits | 200 API credits a month |
| Access to the endpoint tested | Self-serve | Self-serve | Self-serve |
The index scores documentation, reliability, ecosystem and access, which is a different question from job coverage. See the full scores on the Bright Data and Crustdata pages, or compare them directly at Bright Data vs Crustdata.
The bottom line on each provider
- TheirStack. First on open roles at 76.6% (351 of 458). Found at least one role at 48 of the 50 companies.
- Bright Data. Second on open roles at 63.1% (289 of 458). Found at least one role at 41 of the 50 companies.
- Crustdata. Third on open roles at 50.4% (231 of 458). Found at least one role at 36 of the 50 companies.
How to benchmark job postings providers yourself
Start from the companies you care about and read each one's own job board. Greenhouse, Lever, Ashby, SmartRecruiters and Workday all publish a public list of open roles with a posting date for each one. Read the board before and after you query the providers, and keep only the roles listed both times.
Send each provider the same roles, in the form its API takes, and keep everything it returns. Count a hit only when a returned posting is the same job. A matching job id settles it. Otherwise compare title, seniority, specialization and location, and have the cases graded without the grader seeing which provider answered.
Divide by every role on your list, so a provider that returns nothing is counted as missing the role. Then split the score by the age of the role. In this test Crustdata returned 43.2% of roles from the past 7 days and 53.1% of roles 8 to 14 days old, so the age of the role changes the score.
FAQ
Which job postings API returns the most open roles? TheirStack. It returned 76.6% of 458 open roles at 50 companies (351 roles). Bright Data returned 63.1%, Crustdata 50.4%.
Does the order change with the age of the role? The order does not change. TheirStack led every age group, with 82.7% of roles from the past 7 days, then Bright Data, then Crustdata.
Is TheirStack or Bright Data better for job postings data? On this panel, TheirStack returned 76.6% of open roles and Bright Data 63.1%, a gap of 13.5 points. TheirStack found a role at 48 of 50 companies and Bright Data at 41.
How is a job counted as found in this benchmark? Each role came from the company's own job board. A provider scores a hit when a returned posting carries the board's job id, or when model graders that could not see the provider judged it the same job. The score is hits out of all 458 roles. TheirStack scored 76.6%.
More benchmarks
- People search API benchmark 2026: Crustdata, Fiber and Exa
- Authentication API benchmark 2026: Auth0, Firebase, WorkOS, Clerk, Stytch and more
- LLM API benchmark 2026: OpenAI, Claude, Gemini, Mistral, Grok, Groq and DeepSeek
- Transactional email API benchmark 2026: SendGrid, Resend, Postmark, Mailgun and more
- Vector database API benchmark 2026: Pinecone, MongoDB Atlas, Zilliz, Supabase and more
- Text-to-speech API benchmark 2026: ElevenLabs, Google Cloud, Polly, OpenAI and more
- Payments API benchmark 2026: Stripe, Square, Razorpay, Mollie, Braintree, Checkout.com, Adyen
- Maps and geocoding API benchmark 2026: Google Maps Platform, Mapbox, HERE, TomTom and more
- Crustdata vs People Data Labs API benchmark 2026: company and person matching
- E-signature API benchmark 2026: DocuSign, Dropbox Sign, PandaDoc and 4 more
- KYC API benchmark 2026: Stripe Identity, Persona, Plaid, Sumsub and more
- Personal email API benchmark 2026: ContactOut, FullEnrich, Nyne, SignalHire, Wiza and Apollo
- SMS API benchmark 2026: Twilio, Vonage, Telnyx, Plivo, Sinch and more
- Person enrichment API benchmark 2026: Coresignal, Nyne, Crustdata, Fiber and People Data Labs
- Data enrichment API benchmark 2026: Crustdata, People Data Labs, Bright Data and 16 more
- Semantic search API benchmark 2026: Crustdata, Exa and Parallel
- Translation API benchmark 2026: Google, Azure, Amazon, DeepL, Lara, Lokalise and LILT
- Image generation API benchmark 2026: OpenAI, Google Imagen, fal.ai, Replicate and more
- Company enrichment API benchmark 2026: Crustdata, Fiber, Coresignal, People Data Labs and Nyne
- Technographics data benchmark 2026: Crustdata vs BuiltWith vs TheirStack vs Sumble
