How does ml-connector bridge PayPal and TallyPrime if PayPal is cloud and TallyPrime is on a local machine?

ml-connector runs either in the cloud or on-premise and connects to PayPal via HTTPS and OAuth 2.0. To reach TallyPrime, it uses a local agent (a lightweight service running on the same LAN as TallyPrime) that proxies requests to the HTTP server on port 9000. The agent is configured with the TallyPrime host IP and port, so the cloud connector only needs the agent address and credentials.

What happens if TallyPrime is not running when ml-connector tries to post a voucher?

The import request will fail and be retried on the next polling cycle. Since TallyPrime is a desktop application that must be actively running with the target company open, ml-connector treats unavailability as a transient error. The audit trail records all failed attempts, and you can manually trigger a replay once TallyPrime is back online.

Does ml-connector handle PayPal refunds or chargebacks?

Yes. PayPal transaction search includes refunds and chargebacks, which ml-connector syncs as debit or reversal vouchers in TallyPrime. Refunds are mapped to the same ledger account as the original payment so your PayPal clearing account balances correctly. Chargebacks are recorded separately so you can track disputes and recoveries.

TallyPrimePayPal

TallyPrime and PayPal integration

TallyPrime runs day-to-day accounting in South Asia, while PayPal handles worldwide payments and invoicing. Connecting them keeps your TallyPrime ledger in sync with your PayPal transactions. Every payment received or invoice sent through PayPal flows into TallyPrime as a receipt or sales voucher, posted to the right ledger accounts, without manual re-entry. Reconciliation moves from tedious day-to-day to automatic.

How TallyPrime works

TallyPrime is a desktop accounting and ERP application that exposes ledgers, groups (account hierarchies), vouchers (purchase, sales, payment, receipt, orders), and stock items through a local HTTP server on port 9000. The API uses XML or JSON request envelopes controlling operations via TALLYREQUEST field (Import Data for write, Export Data for read). Authentication is optional at the transport layer but requires the target company name and optional credentials if the company is password-protected. No webhooks or change streams exist, so data is read by polling with date-range filters (SVFROMDATE, SVTODATE). The local HTTP server is only LAN-accessible and must be manually enabled in TallyPrime Settings, and the application must be running with the target company open for API calls to succeed.

How PayPal works

PayPal REST APIs cover payments, orders, invoicing, subscriptions, payouts, transaction search, and disputes across sandbox and production environments. All calls use OAuth 2.0 Client Credentials grant with a Base64-encoded CLIENT_ID:CLIENT_SECRET, returning a bearer token (default 8-hour lifetime) cached and reused. PayPal also supports webhooks via HTTPS POST with RSA-SHA256 signature verification to up to 10 registered endpoints per app, with automatic retries up to 25 times over 3 days on non-2xx responses. Key entities include orders, captures, refunds, invoices, transactions, and payouts, with transaction search limited to a 31-day window and a 500-result page maximum.

What moves between them

The main flow is PayPal into TallyPrime. After each payment or invoice in PayPal, ml-connector polls the PayPal transaction search API within a 31-day window, maps each transaction to a TallyPrime ledger (usually a bank or PayPal clearing account), and creates a sales or receipt voucher in TallyPrime with the payment amount and date. Invoices sent through PayPal are also synced as sales vouchers, posting to the appropriate ledger and customer group. Reference data such as customer names and ledger accounts are configured per integration instance.

How ml-connector handles it

ml-connector stores both credential sets encrypted: PayPal OAuth 2.0 CLIENT_ID and CLIENT_SECRET are used to obtain and cache a bearer token (refreshing on 401 or timeout), while TallyPrime requires the local agent IP address, port (default 9000), target company name, and optional credentials. Because TallyPrime has no webhooks and is only LAN-accessible, ml-connector polls PayPal's transaction search API on a configurable schedule (typically daily or per-business-day) within the 31-day transaction window, compares returned transaction IDs against last-seen state, and translates new transactions into TallyPrime Import Data requests. Each PayPal transaction becomes a voucher via the Tally XML envelope with TALLYREQUEST set to Import Data, including the ledger account, customer reference, amount, and date. TallyPrime has no native idempotency, so ml-connector uses transaction ID deduplication to prevent duplicate vouchers from repeated imports. Error responses from either system (PayPal 429 rate limits, TallyPrime LINEERROR blocks, or connection timeouts) are retried with exponential backoff, and full audit trails record every record attempted.

A real-world example

A mid-sized e-commerce or SaaS business in India uses TallyPrime for accounting and GST filing, and PayPal for worldwide customer payments and invoicing. Before the integration, the finance team exported PayPal transaction reports and manually entered each payment into TallyPrime ledgers by date and customer, a task that took several hours each week and introduced transcription errors. The process was especially painful month-end, when reconciling PayPal to TallyPrime required hand-matching dozens of entries. With TallyPrime and PayPal connected, every PayPal transaction automatically flows into TallyPrime as a sales or receipt voucher the same day. The PayPal clearing account ledger is already reconciled, and GST calculations reflect the actual transaction dates and amounts. Month-end close is now a matter of confirming no exceptions rather than re-keying hundreds of entries.

What you can do

Sync PayPal payment transactions into TallyPrime receipt or sales vouchers daily, mapped to the correct ledger accounts and customer groups.
Post PayPal invoices sent to customers into TallyPrime as sales vouchers with ledger detail and customer reference.
Handle OAuth 2.0 token refresh and local TallyPrime HTTP server transport, including local agent bridging and company authentication.
Deduplicate PayPal transaction IDs to prevent duplicate vouchers when imports are replayed or retried.
Maintain a full audit trail of every PayPal transaction attempted, with error replay and reconciliation visibility.

Questions

How does ml-connector bridge PayPal and TallyPrime if PayPal is cloud and TallyPrime is on a local machine?: ml-connector runs either in the cloud or on-premise and connects to PayPal via HTTPS and OAuth 2.0. To reach TallyPrime, it uses a local agent (a lightweight service running on the same LAN as TallyPrime) that proxies requests to the HTTP server on port 9000. The agent is configured with the TallyPrime host IP and port, so the cloud connector only needs the agent address and credentials.
What happens if TallyPrime is not running when ml-connector tries to post a voucher?: The import request will fail and be retried on the next polling cycle. Since TallyPrime is a desktop application that must be actively running with the target company open, ml-connector treats unavailability as a transient error. The audit trail records all failed attempts, and you can manually trigger a replay once TallyPrime is back online.
Does ml-connector handle PayPal refunds or chargebacks?: Yes. PayPal transaction search includes refunds and chargebacks, which ml-connector syncs as debit or reversal vouchers in TallyPrime. Refunds are mapped to the same ledger account as the original payment so your PayPal clearing account balances correctly. Chargebacks are recorded separately so you can track disputes and recoveries.

Connect TallyPrime and PayPal

Free to use. Add your credentials, ping your real systems, and see if we fit.

Get started