BVB-INFO : Free API solution for BVB data | API Gratis pentru BVB 🇷🇴

Free Bucharest Stock Exchange data: scraper, database, API, and a terminal website. No key, no signup, fully open-source and free to use. Toate scraperele/APIurile pentru BVB erau platite, print Tiriac, Tradeville etc. Acesta este un scraper gratis de date JSON. Fara signup sau alte kkturi. Ceri API, primesti, folosesti date instant in proiectul tau.

Terminal: bvb-api.vercel.app · API docs/Documentatie FastAPI: bvb-api.vercel.app/api/docs

---

EN: The Bucharest Stock Exchange publishes plenty of data, but not in a way you can curl. This project fixes that. Every weekday after market close, a scraper walks BVB's own public endpoints, tidies the Romanian-formatted numbers into a proper schema, and loads everything into Postgres. A small FastAPI app then serves it two ways: as a browsable website and as a free JSON API.

RO: BVB-ul tine o tona de date despre companiile locale. Totusi, toate solutiile API pentru a extrage datele acestea sunt platite, cauzand un impas pentru persoanele care doresc sa faca aplicatii/websiteuri fara sa plateasca folosind aceste date. Acest scraper FastAPI este rapid, nelimitat, si up-to-date (cron job cand se inchide marketul la ora 19).

EN: If you've ever wanted to build something on Romanian market data; a screener, a dashboard, a spreadsheet that updates itself; this is the missing ingredient.

RO: Daca ati vrut vreodata sa faceti orice cu datele BVB FARA sa platiti pentru ele si fara sa stati sa consumati tokenuri pe claude code sau pe codex degeaba, mai bine folositi acest screener/dashboard/spreedsheet cu auto update PRIN API!!!

What's in the data

• ~400 instruments — every listed company, fund, and ETF, with profile, ISIN, segment, and share structure
• Daily prices — OHLC, volume, traded value; one row per instrument per session, accumulating history over time
• Valuation metrics — market cap, P/E, P/BV, EPS, dividend yield
• Shareholders — who owns what, in percentages and share counts
• Index constituents — BET and friends
• Company news — the announcement feed, per symbol

The API

Everything is under /api/v1, everything is GET, and everything returns JSON. Responses are cached at the edge, so go ahead and hit it from your cron job — you'll mostly be talking to a CDN, not our database.

# What's in the dataset right now?
curl https://bvb-api.vercel.app/api/v1/

# Banca Transilvania: profile, metrics, shareholders, news, latest price
curl https://bvb-api.vercel.app/api/v1/companies/TLV

# Every company with "energ" in the name
curl "https://bvb-api.vercel.app/api/v1/companies?search=energ"

# Price history for a symbol
curl "https://bvb-api.vercel.app/api/v1/prices?symbol=SNP&date_from=2026-07-01"

# Who's in the BET index?
curl https://bvb-api.vercel.app/api/v1/indices/BET

Endpoint	What you get
/api/v1/	Dataset status — counts, freshness, available indices
/api/v1/companies	Search & filter companies (search, segment, category, limit, offset)
/api/v1/companies/{symbol}	The works: profile + metrics + shareholders + news + latest price
/api/v1/prices	Daily prices (symbol, date_from, date_to, paginated)
/api/v1/indices · /{name}	Index list and constituents
/api/v1/news	Announcements (symbol, paginated)

Interactive docs with a try-it button live at /api/docs.

Run it yourself

Python 3.11+. The scraper works out of the box against a local SQLite file — no server, no config:

python -m venv .venv && source .venv/bin/activate
pip install -r requirements.txt

python -m bvb_scraper.main init-db      # create tables
python -m bvb_scraper.main update-all   # scrape everything (~15 min)

uvicorn bvb_scraper.api:app --reload    # website + API on :8000

That's it — open http://localhost:8000 and you have your own private copy.

Want Postgres instead? Point BVB_DATABASE_URL at it and run alembic upgrade head first. There's also a docker compose up --build that does Postgres + migrate + scrape in one go.

Scraper commands

Command	Does
update-all	Full run: discover → prices → indices → per-company details
update-all --with-filings	…plus download filings per company
update-all --no-details	Prices + indices only (fast)
prices	Just today's price snapshot
company TLV	One company, fetched and persisted
filings TLV	Discover + download one company's filings
discover	List every discoverable symbol

All tunables (request pacing, retries, workers, log level) are BVB_* env vars — see .env.example.

How it stays fresh

A GitHub Actions workflow (scrape.yml) runs the full update every weekday at ~19:20 Bucharest time, after the 18:00 close. The site shows a "DATA AS OF" stamp so you always know what session you're looking at.

How it's put together

crawler/   fetches raw HTML/JSON from bvb.ro (thread-local sessions, retries)
parsers/   turns responses into typed Pydantic models (BeautifulSoup, no regex)
etl/       maps Romanian labels to canonical fields, orchestrates the run
db/        SQLAlchemy 2.0 schema + dialect-agnostic upserts (Postgres/SQLite)
api/       FastAPI: /api/v1 JSON + the Jinja2-rendered frontend
main.py    thin CLI over the pipeline functions

Design choices worth knowing: selectors anchor on stable element IDs (never regex over HTML), one failing company never aborts a run, and the public API reads through a separate read-only database role. The frontend is server-rendered with zero build step — one CSS file, a splash of vanilla JS.

Adding an endpoint is a well-worn path: URL in config.py → fetch function in crawler/ → parse in parsers/html.py → label map in etl/normalize.py → table + upsert in db/ → wire into etl/pipeline.py → test with a saved HTML fixture. The README of your dreams it is not, but git log shows it works.

When something breaks

• Empty results / connection errors — BVB rate-limits sometimes; retries are automatic. Bump BVB_REQUEST_DELAY if it persists.
• "target database is not up to date" — alembic upgrade head.
• Fields suddenly missing — BVB renamed an element ID; update parsers/html.py and etl/normalize.py.
• Chatty logs wanted — BVB_LOG_LEVEL=DEBUG shows every request.

DISCLAIMER

BVB as well as the scraper can lag, gap, or be wrong. Treat it as a convenient market feed, not an exact exact exact replica. Nothing here is investment advice obviously. Don't abuse my rate limits pls (pls!) 🙏🙏🙏🙏

---

For rate limits or broken data problems, please reach out to me at aethrex@proton.me :))) Daca aveti probleme cu coruptie de date sau orice altceva, va rog sa imi scrieti la aethrex@proton.me :)))

from arbust 🇷🇴 with love