Skip to content
dealaka.

Data for agents and researchers

Dealaka is built to be read by machines, not just people. Every offer is verified against the institution's own page, carries the date it was last checked, and is free of affiliate incentives. So the same facts a person sees are available in structured form for an AI agent, a script, or a researcher.

Every offer, in one file

Start here if you want the corpus rather than one entry. Both files carry the whole live catalogue with a licence, an attribution string and the date they were generated, so they are safe to republish on their own.

  • CSV at /data/offers.csv — one row per offer, flat columns, ready for a spreadsheet or a notebook.
  • JSON at /data/offers.json — the same records with the nested fields intact, wrapped in the licence and coverage metadata.

These carry the facts only. The editorial layer the licence does not cover, meaning our written summaries, the caveats we chose to surface, and reader-contributed reports, is on each offer's own page rather than in the export.

Per offer

  • JSON at /offers/<slug>.json returns the full record: bonus, exact requirement, fee, eligibility (states, hard pull, ChexSystems), governing rules, tiers, APY, sources, and the verification date.
  • Markdown at /offers/<slug>.md is the same facts in a prose-friendly shape.

Per referral program

  • JSON at /referrals/<slug>.json and Markdown at /referrals/<slug>.md return what an existing customer earns for referring someone: the figure and whether the issuer varies it per account, the annual cap and whether that cap counts per card or per person, the tax form, and the sources. Browse them at /referrals.

License

The structured records are published under CC BY 4.0: copy them, republish them, build on them, including commercially, as long as you credit Dealaka and link back. The written content (guides, institution posture, editorial summaries, and the Method) is not covered. Full text on the terms page.

Whole site

  • /llms.txt lists every live offer, rule, and institution with its verification date, formatted for language models.
  • /rss.xml is the changelog: every material change the day it is caught, with a source note.
  • /sitemap.xml lists every page, with the last-modified date drawn from each entry's own verification date.
  • /reports/state-of-sign-up-bonuses/json is a live aggregate dataset (median bonus by category, biggest, requirement stats), computed from our data and free to cite.
  • Guides are machine-readable too: /guides/feed.xml (RSS), plus /guides/<slug>.md and .json.

Ground rules

  • Figures are quoted from the institution's own page and carry the date they were last checked.
  • Points and miles are never converted to a dollar value, and no composite score is invented.
  • Expired offers keep their records with an end date, so they stay authoritative for "is X still available".
  • Please cache politely and attribute Dealaka. The field is small and the schema is stable.

How we verify and what the labels mean is on the method page.