Looking for a Karza GST API Alternative

Karza is a well-known Indian verification and KYB data provider, and if you are reading this you are probably holding a renewal quote and wondering whether the arrangement still fits. This post will not tell you what any other company charges or how its product works — we cannot verify that, and you should read their current pricing and documentation on their own site. What it gives you is the procurement method: what you are genuinely paying per useful verification, what to put in writing before you sign again, and what a migration actually costs.

The engineering side of a switch — the technical evaluation and the parallel-run harness that proves the new provider agrees with the old one — is covered in looking for a Surepass GST API alternative. This one is for whoever owns the contract.

Why teams start looking in the first place

Almost nobody goes shopping for a verification API because the data was wrong. They go shopping because of a line on an invoice. The triggers repeat across every finance team we talk to:

• An annual minimum you did not hit. You committed to a volume on a forecast that assumed onboarding growth, and the growth landed somewhere else.
• A renewal uplift. A higher number than last year, no change in what you receive, and no mechanism in the contract that explains how the number was reached.
• A bundle you barely use. Six verification products on the invoice, one in production.
• A quote that only arrives after a sales call. Finance cannot model a cost with no published rate card, and engineering cannot benchmark a service it cannot call.
• Credits that expired. You bought a pack sized for a project, the project slipped a quarter, and the balance went to zero on a date rather than on use.

None of these are complaints about data quality, and that matters: the decision in front of you is a commercial one, and it should be settled with arithmetic rather than a feature comparison.

Working out what you actually paid

Your finance system knows what you spent. It does not know what you bought. Reconstruct the last twelve months line by line, and be strict about the denominator.

In the numerator goes everything that left the account: committed spend whether or not you consumed it, top-up packs, and credits that expired unused. In the denominator goes only verifications that answered a question someone acted on — not calls burnt on GSTINs that failed format or check-digit validation, which you could have rejected offline for nothing, and not duplicate lookups inside a window where the answer could not have changed.

A nominal ₹0.40 became a real ₹1.16, and not one rupee of that gap came from the vendor doing anything wrong. It came from a commitment sized to a forecast, credits with a date on them, and calls nobody filtered. The method is set out in GST API cost per verification; the traps that inflate the numerator are catalogued in hidden costs of a cheap GST API.

A worked example: 40% utilisation versus ₹0.50 a call

Take the same team, and set the expired top-up aside to keep the comparison conservative. Assumptions stated up front: a ₹1,00,000 annual commitment covering 2,50,000 verifications, 1,00,000 calls made, 6,000 malformed and 4,000 duplicates, leaving 90,000 useful verifications.

• Committed plan. ₹1,00,000 ÷ 1,00,000 calls made = ₹1.00 per call. Against useful verifications: ₹1,00,000 ÷ 90,000 = ₹1.11.
• Pay-as-you-go at ₹0.50. Reject the 6,000 malformed GSTINs offline, cache away the 4,000 duplicates, and you buy 90,000 verifications: 90,000 × ₹0.50 = ₹45,000, or ₹0.50 per useful verification.
• Pay-as-you-go at the bulk rate. A single purchase above 2,500 credits is billed at ₹0.40, so 90,000 × ₹0.40 = ₹36,000, or ₹0.40 per useful verification.

That is ₹1,00,000 against ₹36,000 for the same 90,000 answers — ₹64,000 of difference in one year, most of it unconsumed commitment and the rest waste you can remove yourself in an afternoon.

The honest caveat: at full utilisation the committed plan lands at ₹0.40, exactly our bulk rate. A commitment is not inherently a bad deal — it is a bad deal when you miss it, and you find out which one it was twelve months after you had to decide. Other volumes are worked through in what 1,000 GSTIN verifications really cost and lowest cost bulk GSTIN verification.

Unbundling: a KYB suite is a decision, not a default

Verification products are often sold as a suite — GSTIN, PAN, bank account, director records, document checks — because a suite is easier to sell and harder to leave. For a team genuinely running full business onboarding, that can be the right purchase.

The test is consumption: what fraction of last year's contract value did you actually use? Pull the usage report by product. If GSTIN verification is 90% of your call volume and 15% of the contract value, you are funding five products to run one. Make that choice deliberately at renewal rather than inheriting it for a fourth year. Where GST data does and does not cover business verification is set out in KYB verification in India with a GST API.

The questions to put in writing before you renew

Send these in one email, with a reply-by date, to your incumbent and every alternative on your shortlist, including us. Written answers are the contract you are actually signing; the call is not.

None of these are adversarial. A vendor with clean terms answers all eight in one reply and is usually relieved to be asked; a vendor who needs a call to answer question one has told you something about questions two through eight. Storing GSTIN data under the DPDP Act covers what you remain responsible for regardless of which vendor you pick.

What a clean pay-as-you-go arrangement looks like

Most of the eight questions exist because of a pricing structure, not because of a vendor. Per-call pricing with no expiry deletes them. No minimum, and question two is moot. No expiry, and question three is moot. No contract, and questions four and five are moot — there is nothing to renew and nothing to give notice on, so next year's price is the published price and you leave by not buying more credits.

What remains is what should matter: is the data right, is the service up, and is the response deep enough to answer your question in one call. The argument is made at length in using a GST API without a subscription and GST API pricing explained.

The switching cost most teams overestimate

The most common reason a bad contract renews is a belief that migration is a quarter of engineering work. For a verification API it usually is not. You are replacing one outbound HTTP call and the function that maps its response into your model — no state to migrate, no webhook choreography, no re-authentication. Most teams are looking at a day or two of code. The real work is elsewhere:

• The field-by-field diff. Two providers can both be correct and still disagree on spelling, date formats, address structure or how a cancelled registration is labelled. Map every field you consume and decide what your code does when the shape changes; the parallel-run harness in the Surepass alternative post is the tool for it.
• The vendor-master re-run. Every GSTIN you hold was verified against the old source, at a date you may not have stored. At cutover, re-verify the whole master through the new provider and diff it against your stored values. Anything that disagrees on status, legal name or filing recency goes to a human queue before it clears for payment; anything that agrees gets marked verified with a fresh source and timestamp. A queued bulk endpoint makes this a background job rather than a project — POST /api/bulk/jobs takes up to 10,000 GSTINs at a time, and verifying thousands of vendor GSTINs walks through the run.
• The overlap window. Keep both providers live for two to four weeks. The overlap on a pay-as-you-go plan costs a few thousand rupees; finding a field mismatch after you cancelled the old key costs a payment run.

Our terms, held to the same checklist

We build one of these APIs, so treat this as a disclosed interest and check it rather than trusting it. Our answers to our own eight questions:

• Rate card. Published at /pricing: ₹0.50 per verification pay-as-you-go, ₹0.40 once a single credit purchase crosses 2,500 credits.
• Minimum. None. No subscription, no monthly floor, no annual commitment.
• Expiry. Purchased credits do not expire. The 20 free credits released on email verification do expire, 30 days after issue.
• Notice period. None, because there is no contract to give notice on. You stop by not buying more credits.
• Renewal price. There is no renewal event. The published rate is the rate.
• Sandbox. Self-service. Sign up at /register, verify your email, and call real GSTINs with the 20 free credits — no sales call. There are also five free lookups a day at the public search tool, no signup.
• SLA. A published uptime commitment with a stated remedy, and p95 latency under 350ms. Registration details and GSTR filing history come back in a single call, so the denominator in your cost calculation is not halved by a second request. The response shape is at /docs, integration guides at /developers.
• Data and DPDP. Our retention position is written down rather than described on a call.

Where we are not the answer: if you genuinely consume a full KYB suite — bank penny-drops, director graphs, document forensics — a single-purpose GSTIN API does not replace it. Hold us to the same eight questions you send everyone else, and for the broader landscape before you shortlist, read how to choose a GST verification API provider in India and the pricing arithmetic in the cheapest GSTIN API.

Frequently asked questions