SBI TT Rates Archive

This repository serves as an archive containing historical SBI TT rates since 02 July 2020. These rates are one of the important rates required for ITR (Income Tax Return) purposes and are unfortunately not made readily available by RBI (Reserve Bank of India) or SBI (State Bank of India).

This repository is forked from skbly7/sbi-tt-rates-historical

Note

This repository is maintained to provide historical data on SBI TT rates. The rates are collected and stored automatically, and while efforts are made to ensure accuracy, no guarantees can be made regarding the completeness or correctness of the data.

Overview

The repository is structured to maintain the SBI TT rates in a directory format based on the year and month of the rate data. Each rate file is saved with a timestamp to indicate the exact date and time it was retrieved.

Directory Structure

The files are organized in the following directory structure:

├── 2020
│ ├── 07
│ │ ├── 2020-07-02-00:00.pdf
│ │ ├── ...
│ ├── ...
├── 2021
│ ├── 01
│ │ ├── 2021-01-01-00:00.pdf
│ │ ├── ...
│ ├── ...
├── ...

Each file is named in the format YYYY-MM-DD-HH:MM.pdf, where YYYY is the year, MM is the month, DD is the day, and HH:MM is the hour and minute when the SBI TT Rates file was retrieved.

Usage

You can browse the directories to find the historical SBI TT rates for specific dates. The rates are saved as PDF files and can be viewed using any standard PDF viewer.

CSV Rates

In addition to raw PDFs, this repository provides pre-parsed CSV files in the rates/ directory.

rates/
  USD-INR.csv        ← one file per currency pair
  EUR-INR.csv
  GBP-INR.csv
  ...
  _metadata.csv      ← one row per processed PDF (audit trail)

Each CSV has the schema:

date,tt_buying,tt_selling,bill_buying,bill_selling,card_buying,card_selling,cn_buying,cn_selling,publication_only,source_file

• date: YYYY-MM-DD
• Rate columns: decimal strings; empty string means SBI did not publish that rate for that day
• JPY, THB, and KRW rates are stored per-unit (divided by 100 from the published per-100-unit values)
• publication_only: true for KRW and TRY (flagged as "for publication purpose only" in SBI PDFs)
• source_file: relative path to the source PDF in this repo

Scripts

Requires Python ≥ 3.11 and uv.

# Parse all historical PDFs and generate initial CSVs
uv run python scripts/backfill.py

# Parse a single PDF and append to CSVs
uv run python scripts/daily_update.py 2026/06/2026-06-20-14:00.pdf

# Validate CSV integrity
uv run python scripts/validate.py

Automation

This repository is automatically updated twice daily using a GitHub Action. The action downloads the latest SBI TT rates, commits the PDF, then parses it and updates the rates/ CSVs in a second commit. This ensures that both the raw PDFs and the derived CSVs stay current.

Contributions

Contributions are welcome!
If you have any improvements or suggestions, feel free to open an issue or submit a pull request.