Use the download hub. tigzig.com/downloads is the page for people, and api.tigzig.com/downloads is the JSON index for agents that links to every file. Free, no signup, no API key.
For your AI agent, paste this: "Go to api.tigzig.com and list all downloads".
What is in it, as of 11 October 2026: 46 datasets and 172 files across five apps.
- MFPRO (5 datasets, 17 files): daily NAV for every Indian mutual fund scheme since 2006, the latest NAV list, scheme details (fund size, fund manager, TER), a data quality file, and monthly portfolio holdings of 24 Indian equity funds.
- SIF (3 datasets, 6 files): Specialized Investment Funds - NAV history, the scheme master and quarterly average AUM.
- TREMOR (15 datasets, 55 files): macro indicators, market prices, US bank, credit union and household credit, bank lending to non-bank financial firms, India bank credit, insurance investments, the Trump trade and annual disclosure tables, and all tables in one database file.
- VIGIL (11 datasets, 57 files): India company signals - credit ratings, insider trading, promoter pledges, takeovers, encumbrance, surveillance flags, related-party transactions and BRSR ESG data, plus all tables in one SQLite file.
- CRICKET (12 datasets, 37 files): ball-by-ball deliveries for every format and for the IPL, matches, match players and the player register, plus all tables in one database file.
Each app has its own index for agents: api.tigzig.com/downloads/mf, /sif, /tremor, /vigil and /cricket. Each dataset entry names its source, the date of its latest data and how often it updates (for example, NAV history refreshes 3 times a day and scheme details 4 times a day, while fund holdings are monthly).
Formats. Every table comes as CSV, and many also as Parquet, TSV or SQLite. Macro and markets, India company filings and cricket also publish their main tables as one DuckDB or SQLite file.
Parquet files can be queried in place. The server answers HTTP range requests, so DuckDB, Polars, pyarrow or a browser reads only the parts of the file a query needs: SELECT count(*) FROM read_parquet('https://api.tigzig.com/mf/v1/download/amfi_nav_master.parquet'); More on that: SQL on a parquet file from a browser.
Whether to download or call the API is its own question: can I download the whole dataset instead of calling the API? Release note: https://www.tigzig.com/post/tigzig-downloads-hub-46-datasets-oct2026.
Cricket data comes from Cricsheet (cricsheet.org) under the ODC-BY 1.0 licence. TigZig is not affiliated with or endorsed by Cricsheet.
Building something like this? How I work covers the rates, the availability and what I take on.