Open Data
India Credit Card Facts — a free, dated dataset
The core facts behind CardAdvisor's catalogue, published as clean JSON and CSV for researchers, journalists and AI tools. Independent, no paid placements — free to cite with attribution.
Download
Full dataset incl. per-category reward rows with caps, one point value per row, and perk types.
Flat, card-level table (one row per card) for spreadsheets and quick analysis.
Both endpoints rebuild from the live catalogue at most hourly, and may be served from cache for up to a day after that. Cite the asOf inside the file you downloaded, not the date shown above. This page and the files are cached independently, so the two can represent snapshots a day or two apart — the file’s own asOf is the one that describes the data you actually hold. Programmatic access is welcome — please cache responses and don’t hammer the endpoints. For a single card, the same facts are addressable through the public API (no key needed); for anything wider, these files are the way — that is what they are for. One practical note: our bot protection challenges a few default library user-agents (Python’s urllib among them), so send a descriptive User-Agent that names your project — requests, curl, browsers and AI crawlers all pass as-is.
Change history
The complete changelog behind the changes tracker: every tracked, source-verified change to active Indian credit cards — 1,800 dated events across 280 cards, back to 2006 — with each event’s direction, before/after values and verification tier.
Counting rule: this dataset counts per affected card — a bank-wide revision that touches twenty cards appears as twenty events, each carrying its own card and before/after values. Counted once per distinct change instead (the way the changes tracker and our devaluation study count), the same record holds 792 changes. Both countings are correct; cite whichever answers your question, and say which you used.
Full changelog incl. field dictionary — one record per change event, newest first.
Flat, event-level table (one row per change) for spreadsheets and quick analysis.
Prefer push over pull? The same record is syndicated as an RSS feed (one item per distinct bank act) and an effective-dates calendar (upcoming changes only, each on the day it takes effect — subscribe from any calendar app).
Writing about one card? Every card’s timeline is also embeddable — an “Embed this timeline” snippet sits under the Value Timeline on each card page (for example, the HDFC Infinia timeline). Facts only, attribution built in, free to use.
Points transfer graph
The reward-currency network behind the transfer explorer: 58 currencies and programmes connected by 161 published routes — numeric ratios, tier gates, caps and official sources. (What each currency is worth in rupees is our valuation model — it is on the explorer, not in the dataset.)
Lounge entitlements
What every card on the lounge finder actually opens — 170 cards: domestic, international and railway visits a year, the per-quarter ceiling where one binds, the spend gate in the issuer's own terms, the programme behind the door, guest rules, and announced changes that are not in effect yet. Airport and railway are separate columns and are never summed — 10 of the cards with a railway entitlement have no airport lounge at all. (What a visit is worth is our model, and is not here.)
Point valuations
What one point is realistically worth on every point-earning card — one value each, in paisa, on the default redemption route, for 237 cards across 21 programmes. Read it alongside the reward rates above: a rate without its point value overstates returns several-fold. The route-by-route breakdown (cash, voucher, travel portal, transfer) is our valuation model and lives on the value calculator, not in the dataset.
The catalogue, honestly
A rate table flatters every card it describes. These are the counts a rate table leaves out — computed from the same files as everything above, re-checked on every data edit:
- 352 of the 879 earn rules in the catalogue stop paying at a monthly or annual cap.
- 1,492 card-category combinations pay exactly zero: 262 of 280 active cards exclude at least one spend category outright — most often fuel (236 cards), rent (200 cards) and wallet loads (182 cards).
- 197 earn rules pay only at named merchants or through the bank’s own portal.
- 41 cards cannot be applied for today (15 paused, 26 invite-only) — we label them instead of hiding them.
- 9 closed-audience cards (defence, CAs, and similar) never enter our recommendations — the profiler cannot verify membership, so we refuse to guess.
- 50 cards are lifetime free; of the 217 that charge a fee, 149 publish a spend-based waiver.
Devaluation Index
The ranked edition of the change record: the most-devalued cards and the issuer league table, as published in the October 2026 edition. Unlike the live datasets above, this file is an edition — computed over the 24 months to 2026-10-16, then frozen. When the record moves, the next dated edition carries the delta. The undated file always holds the newest edition, and every edition also keeps its own dated file, so a number you cite from one will never change under you.
Licence & how to cite
Released under CC BY 4.0. You’re free to reuse and republish these figures, including commercially, with attribution to CardAdvisor and a link back to this page. We’re happy to prepare a custom cut or a specific breakdown for a story — just ask.
Want to check the data before you use it? We publish a 30-card provenance audit — every material field with the issuer page it came from, the date we retrieved it, and whether that source corroborated it. Including the fields where it did not.
Citing this in something that has to stay checkable? The endpoints on this page are live — they serve today’s catalogue and different numbers next month. Cite a frozen release instead: a dated snapshot at a URL whose contents never change. Latest is v2026-09-17 (3 releases).
Archived and citable by DOI. This dataset is deposited on Zenodo, which mints a permanent identifier and indexes it through DataCite and OpenAIRE — so a citation resolves whether or not this site is still here. 10.5281/zenodo.21966088 (that is the DOI for release v2026-08-16; cite 10.5281/zenodo.21947844 if you mean the dataset rather than one release).
Copy the HTML version to credit the dataset with a working link; the plain-text version is for print, decks and PDFs.
Source: <a href="https://cardadvisor.in/data">India Credit Card Facts — open dataset</a>, CardAdvisor (CC BY 4.0).
What’s in it
One record per active card. We publish only verifiable facts — fees, eligibility, real reward rates with their caps, one point value per reward row, and the perk types a card carries — never our editorial ratings, our rupee valuations of perks or welcome bonuses, or recommendations. Those are the model: it is published on the site and licensed to businesses (see /partnerships); it is not served through the public API to a free key. Reward rates are the real cash-equivalent earn and deliberately exclude bank-portal accelerators (which apply only to a slice of spend). Schema v3 (16 August 2026) removed the model columns that earlier releases carried — see the release index for the erratum.
| Field | Meaning |
|---|---|
cardKey | Stable slug; the card detail page is /cards/{cardKey}. |
name | Official card name. |
issuer | Issuing bank/network (enum). |
network | Card network (VISA / MASTERCARD / RUPAY / AMEX). |
tier | ENTRY / MID / PREMIUM / SUPER_PREMIUM. |
annualFeeRs | Recurring annual fee in ₹ (0 if none; null if the issuer has not disclosed it). |
joiningFeeRs | One-time joining fee in ₹ (0 if none; null if the issuer has not disclosed it). |
lifetimeFree | True only if genuinely no annual fee AND no spend-based waiver condition. |
feeWaiverSpendRs | Annual spend in ₹ that waives the annual fee; null if not waivable. |
forexMarkupPct | Foreign-currency markup as a percentage (before 18% GST); null if unpublished. |
aprMonthlyPct | Monthly finance charge (APR) as a percentage; null if unpublished. |
minSalaryRs | Minimum income the issuer states, in ₹ a month: a yearly figure is divided by 12 and rounded to the rupee. null if unpublished. |
minAgeYears | Minimum applicant age; null if unpublished. |
maxAgeYears | Maximum applicant age; null if unpublished. |
rewardRateMinPct | Base earn rate as a % of spend (real cash value, excludes portal accelerators). |
rewardRateMaxPct | Best category earn rate as a % of spend (real cash value, excludes portal accelerators). |
lastVerified | Date this card record last CHANGED (yyyy-mm-dd) — a change stamp set on any edit, NOT a claim that the record was re-checked against the issuer that day. |
url | Canonical card detail page. |
rewards | Per-category earn rows: category, rewardType, ratePct (EFFECTIVE cash return as a % of spend, i.e. points-per-₹1 × pointValuePaisa — directly comparable across points and cashback cards), monthlyCapRs, pointValuePaisa, merchants (scope; portal/co-brand rows are marked). Sorted by category, then merchants, with the unscoped row first. |
perks | Perks the card carries, one row per perk, sorted by perkType (INSURANCE, LOUNGE_ACCESS, …). Our rupee valuation of each perk is editorial and is not part of the open dataset. |
Built on this data
Our dated research studies draw on the same catalogue:
- The Devaluation Report (2026) — 24 months of tracked changes measured: 4.5 cuts for every improvement (each bank decision counted once), and the other numbers behind it.
- The Real Point-Value Index (2026) — what a reward point from every Indian card is honestly worth in rupees.
- The State of Indian Credit Cards 2026 — our analysis of 280 cards: lifetime-free reality, forex spread, headline-vs-real rewards.