ქართული
English
Most "BOG vs TBC vs Credo" comparisons stop at three names because that's what Georgian e-commerce looked like a few years ago. It isn't anymore. Georgian merchants today can choose from more than a dozen live payment gateways — bank-owned, fintech, and cross-border — covering cards, installment loans, buy-now-pay-later, QR, open banking, and crypto. This guide covers all of them: what each one actually costs, how they differ underneath the marketing pages, and how to pick the right one for your business.
Why this comparison had to grow
Georgia's e-commerce market generated an estimated $1.1 billion in revenue in 2025, yet only around 31% of Georgian internet users have ever bought something online — compared with roughly 78% across the EU. That's a wide gap, and it's closing: as more Georgian businesses move online, the checkout options they offer — cards, installments, QR, open banking — increasingly decide who gets the sale.
Against that backdrop, a comparison that only covers BOG, TBC, and Credo misses most of the market. Quickpay's own gateway catalog currently includes BOG, TBC, Credo, Keepz, Flitt, UniPay, Liberty Card, Moka, Fastoo, ElectraPay, Paysera, PayPal, and CityPay — 13 live options across cards, installments, BNPL, open banking, crypto, and tokenized recurring billing. This post walks through all of them.
The full landscape at a glance
| Gateway | Underlying Bank / Provider | Card Payments | Installments / BNPL | Other Methods | Settlement | Typical Fee* | Best For |
|---|---|---|---|---|---|---|---|
| BOG Card / Installment | Bank of Georgia | ✓ | ✓ (loan, 45–15,000 GEL) + zero-interest BNPL | Apple Pay, Google Pay | T+1 | ~2.0–2.5% domestic, ~2.8–3.2% intl.* | Broadest reach, highest brand trust |
| TBC Card / Installment | TBC Bank | ✓ | ✓ (campaign-based ganvadeba) | Apple Pay, Google Pay | T+1 (incl. weekends) | ~2% domestic, ~2.5% intl.* | Strong brand, higher-ticket installments |
| Credo Installment | Credo Bank | ✗ (installments only) | ✓ (up to 12,000 GEL, 0% options) | Via Keepz QR (Da-Ka-Vi) | T+2 | Installment-only, quote-based | Simplest API, installment-first stores |
| Keepz | Credo Bank (acquiring) | ✓ | ✓ (via Credo) | QR, NFC tap-to-phone, crypto (via CityPay), open banking | Varies | Not publicly published | In-person/QR, crypto-curious, split payouts |
| Flitt | TBC Bank Group | ✓ | ✓ (50–5,000 GEL) + zero-interest BNPL | Open banking, Apple/Google Pay, QR | Varies | 2.0–2.2% domestic (tiered), 2.5% intl.* | Transparent tiered pricing, cross-border (GE/UZ/MD/AM) |
| UniPay | Liberty Bank (acquiring) | ✓ | – | Apple Pay, Google Pay, Amex, multi-currency (GEL/USD/EUR) | Varies | Quote-based | Billing in multiple currencies (GEL/USD/EUR) |
| Liberty Card | Liberty Bank (direct, TXPG) | ✓ | – | Apple Pay, Google Pay, tokenized recurring | Varies | Quote-based | Direct Liberty Bank integration, subscription billing |
UniPay and Liberty Card are easy to mix up: both ultimately settle through Liberty Bank's card acquiring. UniPay is a payment service provider layered on top of that acquiring (its own dashboard, its own onboarding), while Liberty Card is a direct integration against Liberty's own TXPG gateway — no PSP layer in between. If you already bank with Liberty, Liberty Card is the more direct route; if you want a PSP-style dashboard and Amex support, UniPay is built for that.
Other live options worth knowing
| Provider | What it is |
|---|---|
| Moka (United Payment) | Card acquiring from Moka United, a fintech operating in Turkey, the UK, Azerbaijan, Georgia, Uzbekistan, and Germany — virtual POS, physical POS, and card solutions. |
| Fastoo | Georgian card-payment platform. |
| ElectraPay | Georgian card-acquiring option. |
| Paysera | Lithuanian fintech that opened Georgia's first digital bank in 2023 (licence granted late 2022), with SEPA and SEPA Instant EUR transfers alongside GEL accounts — a strong fit for merchants billing in EUR or serving European customers. |
| PayPal | Global wallet + card acceptance, USD-denominated — useful for international/diaspora customers. |
| CityPay | Standalone cryptocurrency payment acceptance. |
| Bank transfer / Cash on delivery | No-gateway fallback options for manual or COD orders — always free, no module subscription required. |
Deep dive: the three bank gateways
Bank of Georgia (BOG) remains the most widely used card gateway in the country. Its API uses OAuth 2.0 with short-lived JWT bearer tokens, and it ships a drop-in JavaScript calculator widget for installments, which is one reason it tends to integrate faster than TBC in practice. BOG actually runs two separate consumer-credit products worth telling apart: BOG Installment, a standard loan (45–15,000 GEL, 3–48 months, with both 0%-interest "ZERO" and interest-bearing "STANDARD" plans depending on what the bank configures per merchant), and BOG Part-by-Part, a separate zero-interest, short-term BNPL product aimed at smaller baskets. Loan applicants must be Georgian citizens and pass a credit bureau check; the installment system is also GEL-only and Georgian-language only today. TBC Bank matches BOG on brand strength and offers its own online installment system built around merchant-specific "campaigns" rather than one fixed product. Settlement reportedly lands funds the next day, including weekends. The catch most comparisons skip: TBC doesn't hand out production credentials on signup. Yourclient_id/client_secret are only generated after TBC's back office validates and activates your merchant account, and going live requires your site to publish visible Terms & Conditions and a Return Policy, plus completed test transactions — an extra step that a generic "3–5 day setup" figure doesn't capture.
Credo Bank is the outlier: it doesn't accept cards at all — it's purely an installment lender, letting customers borrow up to 12,000 GEL through the MyCredo app, with 0%-interest offers available on some purchases. Its API is also the simplest of the three: a single REST endpoint secured with a shared-secret MD5 request hash, no OAuth flow required. (Credo's own integration PDF warns that the caller's IP has to be whitelisted — in practice that requirement turns out to be outdated, and requests work fine from any server IP. Worth knowing before you spend a support ticket chasing it.) That simplicity is a real advantage for merchants who only want installment checkout and nothing else.
Deep dive: fintech and alternative gateways
Keepz is less a "gateway" than a payment orchestration layer. It was the first fintech in Georgia to hold an Open Banking licence and among the first to bring cryptocurrency payments into everyday retail (via its CityPay integration) — its QR standard is reportedly used by more than 20,000 merchants, letting customers pay via card, bank-to-bank transfer, NFC tap-to-phone, or crypto wallet with no POS terminal and no app download. Card acquiring behind Keepz runs through Credo Bank, and Keepz QR terminals can also trigger Credo's "Da-Ka-Vi" installment option — so Keepz effectively gives Credo's installment product a card- and QR-based front door it doesn't have on its own. The trade-off is integration weight: every Keepz eCommerce request and response is individually wrapped in AES-256-CBC + RSA-OAEP hybrid encryption, which is meaningfully more complex to build against than a standard REST+OAuth API. Flitt is TBC Bank Group's separate, internationally-oriented card gateway, built for merchants registered in Georgia, Uzbekistan, Moldova, or Armenia, with PCI DSS Level 1 certification and prebuilt plugins for several major e-commerce platforms. It's also one of the few providers in this whole list that publishes an actual fee schedule: around 2.2% per transaction up to 10,000 GEL in monthly volume, dropping toward 2.0% above that, with custom enterprise pricing at higher volumes, plus roughly 2.5% on international cards. Flitt also carries its own installment product (50–5,000 GEL) and a zero-interest 4-part BNPL option. UniPay, founded in 2012 and running on Liberty Bank's acquiring, is one of the longer-standing PSPs in Georgia. It supports Visa, Mastercard, and American Express plus Apple/Google Pay, and — unlike every other gateway in this comparison — processes GEL, USD, and EUR directly, with multi-currency accounts available on request. It doesn't offer installments, but that native multi-currency support is a genuine differentiator for merchants who bill outside GEL. (Amex isn't exclusive to UniPay, for what it's worth — BOG and Flitt both list Amex acceptance too, per their own integration docs.) Liberty Card connects directly to Liberty Bank's own TXPG gateway rather than through a PSP layer, and it's the newest addition to Quickpay's catalog. It supports standard card capture plus Apple Pay and Google Pay, EMV 3-D Secure without redirecting to a hosted page, and tokenized merchant- and customer-initiated recurring billing (MIT/CIT) — useful if you're building subscription or installment-style repeat billing without going through a separate subscriptions add-on. It's also one of only two providers in this comparison (with BOG) whose API supports card-to-card and P2P credit transfer operations. Moka (United Payment), Fastoo, and ElectraPay are newer, smaller card-acquiring entrants; Paysera stands out for cross-border EUR business given its Georgian digital banking licence and direct SEPA access.Installments and BNPL, side by side
Georgian shoppers use installment loans heavily for higher-ticket purchases (electronics, furniture, appliances), so this deserves its own comparison:
| Product | Provider | Amount Range | Term | Interest |
|---|---|---|---|---|
| BOG Installment | Bank of Georgia | 45–15,000 GEL | 3–48 months | 0% or standard, set per merchant |
| BOG Part-by-Part | Bank of Georgia | Smaller baskets | Short-term split-pay | 0% |
| TBC Installment | TBC Bank | Set per merchant campaign | Set per campaign | Varies by campaign |
| Credo Installment | Credo Bank | Up to 12,000 GEL | Varies | 0% options available |
| Flitt Installment | TBC Bank (via Flitt) | 50–5,000 GEL | Varies | Standard |
| Flitt BNPL | TBC Bank (via Flitt) | 50–5,000 GEL | 4 parts | 0% |
The part most comparisons skip: integration complexity
Fee percentages get all the attention, but for anyone weighing "integrate directly" against "use an aggregator," authentication complexity is the more honest cost driver:
- BOG — OAuth 2.0 client-credentials flow with JWT bearer tokens; ships a JS widget with a ready-made installment calculator, which shortcuts frontend work.
- TBC — OAuth 2.0 plus an
apikeyheader, gated behind back-office merchant activation beforeclient_id/client_secretare even issued. - Credo — No OAuth at all: a shared-secret MD5 request hash, nothing else. The lightest integration in this list, though its own documentation has enough small inconsistencies (an endpoint path that's documented wrong, a worked hash example that doesn't actually compute) that you should log the raw response and verify against production rather than trust the PDF line-by-line.
- Keepz — Every request and response is individually encrypted with AES-256-CBC and RSA-OAEP, on top of standard REST — by far the heaviest lift, though it buys access to the broadest feature set (cards, open banking, crypto, subscriptions, split settlements).
- Liberty Card — Supports both a modern JSON REST API and a legacy XML
/Execendpoint secured with a client certificate you generate yourself — two payload formats to choose between, which is unusual among Georgian gateways.
Security and regulation: what actually matters
Every provider above is either a licensed bank or a registered Payment Service Provider under the National Bank of Georgia (NBG), which oversees PSP registration and — since a 2023 regulatory update — open banking services specifically. That oversight isn't just theoretical: in 2025 the NBG fined PSPs CPay 14,000 GEL and PayUnicard 39,000 GEL for insufficient AML/KYC controls — a reminder to check that any provider you integrate is currently registered and in good standing, not just live. Beyond licensing, look for PCI DSS compliance (BOG, TBC, Credo, Flitt, and Quickpay itself all advertise it) and 3-D Secure on card transactions as baseline requirements, not differentiators.
How to choose
- Simple online store, cards only: BOG or TBC direct — the two most recognized brands with the widest domestic card coverage.
- High-ticket goods (electronics, furniture, appliances): Prioritize installment coverage — BOG, TBC, or Credo — over card-only gateways; Georgian shoppers convert noticeably better with a financing option visible at checkout.
- In-person, QR, or crypto-curious retail: Keepz — the only option here with a no-POS QR standard and native crypto acceptance.
- Selling into Uzbekistan, Moldova, or Armenia, or wanting transparent published pricing: Flitt.
- Need to bill in USD or EUR directly, not just GEL: UniPay's the only gateway here with native multi-currency accounts.
- Subscription or recurring billing on Liberty Bank: Liberty Card's tokenized MIT/CIT support handles this without a separate add-on.
- Billing in EUR to European customers: Paysera, via direct SEPA access.
- Don't want to pick just one: This is the case an aggregator is built for — see below.
What it actually costs to go through an aggregator
Since the pitch behind a comparison like this one is often "use an aggregator instead," it's worth being specific about what that costs rather than leaving it at "one integration." Quickpay doesn't take a cut of transaction volume — the platform fee is 0% on payment volume. Instead, each gateway is a separate module: every gateway includes a permanent free tier for low-volume use, and paid tiers scale up by monthly transaction count (Starter → Pro → Unlimited), with a discount for paying 3, 6, or 12 months at once, or a one-time lifetime option. Current pricing per gateway is listed on the modules page and can change, so treat any number quoted elsewhere as a starting point, not a final one. On top of the module fee, you still pay whatever the underlying bank charges per transaction — the aggregator doesn't remove that cost, it removes the integration and maintenance cost of connecting to each bank's API (and each bank's auth model) individually.
კომენტარები (0)
იყავით პირველი, ვინც კომენტარს დატოვებს.
დატოვეთ კომენტარი