newTejHQ MCP is live: backtest Indian market data with your own AI agent
TEJHQ

data terms · updated 2026-11-08

What you can do with the data

Plain-language terms for TejHQ data, from the API and the parquet dataset. Questions: support@tejhq.dev.

Summary

  • The data is built from end-of-day bhavcopy and corporate-action files that NSE and BSE publish openly on their websites.
  • TejHQ holds no licence from NSE, BSE or NSE Indices, and has no written clarification from them. We cannot grant rights on the exchanges' behalf.
  • The MIT licence covers our code (the tej-bazaar pipeline and the Python SDK). It does not cover the data.
  • These terms cover what TejHQ adds: the cleaning, adjustment, packaging and the API service. The exchanges' own terms apply independently.
  • Personal use is free: research, backtesting, paper trading, your own trading decisions, local storage and private dashboards.
  • Using the data in a product, for clients, or inside a business is commercial use and needs the Pro plan.
  • Publishing or reselling the raw data is not permitted under these terms.

Where does the data come from?

Daily bhavcopy (end-of-day prices and volumes) and corporate-action files published openly by NSE and BSE. TejHQ downloads them each trading day, cleans them, adjusts prices for splits, bonuses and dividends, and publishes the result as an API and as parquet files. NSE coverage starts 2010-01-04 and BSE coverage starts 2024-07-08.

F&O data comes from NSE's daily equity derivatives bhavcopy, from 2010-01-04. It is served only through the API to Retail and Pro plans and is not part of the free parquet dataset.

Universe files (for example nse_liquid) are TejHQ's own lists ranked by trailing turnover. They are not official index constituents, and TejHQ does not publish index values or other NSE Indices products.

Does TejHQ have a licence from NSE or BSE?

No. TejHQ has no data licence or agreement with NSE, BSE, NSE Data & Analytics or NSE Indices, and has not received written clarification from them about the uses described here. If you need rights confirmed by an exchange, refer to that exchange's terms of use or contact NSE Data & Analytics for a formal licence.

What does the MIT licence cover?

Only the code: the open-source tej-bazaar pipeline and the tejhq Python SDK. The MIT licence does not cover the market data, which comes from the exchanges.

What counts as personal use?

Use by one individual for their own purposes, not for anyone else and not inside a business. This includes:

Historical research and analysis. Backtesting strategies. Paper trading and forward testing. Scans and signals for your own trading decisions. Calculating indicators. Training your own models. Private alerts. Private dashboards and spreadsheets that only you can access. Academic study, including publishing your findings as charts, tables and summary statistics.

Personal use is free with the keyless API, a free key, or the parquet dataset. Retail (₹599/month) is also for personal use, with higher limits and paid endpoints.

Can I fetch the data automatically, backfill history and schedule daily updates?

Yes, within the rate limits. Keyless requests are limited to 120 a minute per IP. A free key gets 1,000 calls a day and 300 a minute. Retail gets 10,000 a day, Pro 500,000 a day.

For full history, download the parquet files from HuggingFace (tejhq/indian-markets) instead of calling the API per symbol. There are no limits on parquet downloads.

Use only the endpoints documented at https://api.tejhq.dev/openapi.yaml. Requests to undocumented paths and abusive traffic may be blocked without notice.

Can I store the data locally and keep it?

Yes. For personal use you may store the data in a local database or files (parquet, CSV, spreadsheets), make private backups, and keep it indefinitely. There is no deletion requirement if you stop using TejHQ or cancel a paid plan.

Who owns the results I derive from the data?

You do. Indicator values, signals, backtest trades, equity curves, statistics, models and reports are yours to keep and use, including after the dataset or service changes.

What counts as commercial use?

Any use for or inside a business, or for other people. For example: an app, website, bot or API that shows the data or anything derived from it to users. Work delivered to clients. Use by a fund, PMS, RIA, broker or advisory firm. Charts or data in a paid newsletter, course or community. Internal tools at a company.

Commercial use needs the Pro plan (₹4,999/month). For several people, email support@tejhq.dev. A paid plan covers TejHQ's side. If you display or redistribute exchange data to others, the exchanges' own display and redistribution terms may also apply, and complying with them is your responsibility.

If you cancel Pro, you may keep the data you stored while subscribed, for your own records, subject to the exchanges' terms.

Can I redistribute or publish the data?

Not the raw data. Under these terms you may not publish, mirror, resell or serve the dataset or API responses as data, on any plan. Publishing your own analysis is fine: charts, tables, summary statistics and research results that do not let someone reconstruct the dataset.

Do I need to give attribution?

Not for personal use. If you publish research or charts made with the data, a credit is appreciated, for example: Data: TejHQ (tejhq.dev), from NSE/BSE public bhavcopy.

Is the data guaranteed to be accurate?

No. The data is provided as is, without warranty. Exchange files sometimes contain errors, and corporate-action adjustment can be wrong. Check the data before relying on it for trading decisions. Nothing TejHQ publishes is investment advice. Please report errors to support@tejhq.dev or as a GitHub issue.

When is the data updated, and is there an SLA?

Every trading day. Data is usually ready by the next morning IST. There is no SLA on free plans. GET https://api.tejhq.dev/v1/status shows the latest trading date loaded.

Can these terms change?

Yes. The date at the top of this page shows the latest version. Data you fetched under earlier terms can still be used as those terms allowed. Last updated 2026-11-08.