API Documentation

Authentication

Send your API key with every request, either way:

x-api-key: tvd_your_key_here
# or
Authorization: Bearer tvd_your_key_here

Keys are issued on subscription and deactivated on cancellation. One key per subscription — don't share it. Two daily limits apply (5,000 requests, 2,500 distinct tickers) — see Limits.

GET/v1/fundamentals

Returns, for each concept, the most recent fiscal period whose 10-K had been filed on or before as_of — the point-in-time view. Note the boundary is inclusive: a 10-K filed on as_of itself counts as knowable that day (filings often land after market close — use the prior day if you need strict before-the-open semantics).

Query parameters

Response fields

Example

curl "https://tradevodata.com/v1/fundamentals?ticker=AAPL&as_of=2024-06-30&concept=Revenue" \
  -H "x-api-key: tvd_your_key_here"

{
  "ticker": "AAPL",
  "as_of": "2024-06-30",
  "data_through": "2026-07-24",
  "count": 1,
  "fundamentals": [{
    "concept": "Revenue",
    "fiscal_year": 2023,
    "period_end": "2023-09-30",
    "filed": "2023-11-03",
    "lag_days": 34,
    "value": 383285000000,
    "latest_value": 383285000000,
    "restated": false,
    "qa_status": "clean"
  }]
}

Note the answer is FY2023, not FY2024 — on 2024-06-30 the FY2024 10-K had not been filed yet. That is the whole product.

Python

The official client. as_of is a requiredargument on every query — there is no way to ask it for “Apple's revenue”, only for what was knowable on a given date, which is how lookahead bias stops being something you can write by accident. Zero dependencies; pandas is optional.

pip install tradevodata

import tradevodata as tv

df = tv.sample()                    # free 40-company sample — no API key needed

client = tv.Client(api_key="tvd_...")           # or set TRADEVODATA_API_KEY
client.fundamentals("AAPL", as_of="2024-06-30") # -> FY2023
client.snapshot(as_of="2024-06-30", concept="Revenue", to_pandas=True)

Source: github.com/christianpichichero-max/tradevodata-py · PyPI

Note: as of June 2024, the newest knowable Apple annual revenue is FY2023 — FY2024 wasn't filed until 2024-11-01. That's the product.

Python (plain requests)

import requests

r = requests.get(
    "https://tradevodata.com/v1/fundamentals",
    params={"ticker": "AAPL", "as_of": "2024-06-30"},
    headers={"x-api-key": "tvd_your_key_here"},
)
for f in r.json()["fundamentals"]:
    print(f["concept"], f["value"], "known since", f["filed"])

R

library(httr)

r <- GET(
  "https://tradevodata.com/v1/fundamentals",
  query = list(ticker = "AAPL", as_of = "2024-06-30"),
  add_headers(`x-api-key` = "tvd_your_key_here")
)
for (f in content(r)$fundamentals) {
  cat(f$concept, f$value, "known since", f$filed, "\n")
}

Errors

An empty result with count: 0 and a note means the ticker exists but nothing had been filed on or before your as_of.

GET/v1/health

Public liveness check (no key needed): returns { ok, rows, db_ms }.

Limits

Two independent daily limits per key. Both reset at 00:00 UTC, and every 429 carries a Retry-After header with the seconds until reset.

The honest why: the dataset is the product. A real backtest touches a few hundred names; a full-universe scrape is all 5,193. The ticker cap sits between those two on purpose — it's an anti-bulk-extraction wall, not a revenue lever. For whole-universe work, don't fight the cap — use the bulk endpoints below (included in your plan).

Bulk & cross-section

For whole-universe work — cross-sectional screens, factor backtests — don't loop the per-ticker endpoint. Both are included in your $49/mo plan.

curl -H "x-api-key: tvd_..." https://tradevodata.com/v1/download -o tradevodata.csv.gz
curl -H "x-api-key: tvd_..." "https://tradevodata.com/v1/snapshot?as_of=2024-06-30&concept=Revenue"