Overview

The Feed API ingests 32 news sources every 60 seconds — including Bloomberg, Cointelegraph, WSJ, Financial Times, CoinDesk, and more — extracts matching stock, crypto, and commodity tickers from every article, and serves them through a simple REST API designed for algorithmic trading bots.

Every article includes a sentiment score (-1.0 to 1.0) and a list of tickers mentioned, so you can filter by symbol and feed the results directly into your trading strategy.

# Quick test — this works right now
curl https://api.dev.thefeedapi.com/v1/news?symbol=BTC&limit=3 | jq

Authentication

Required All API requests require an API key. Sign up free at thefeedapi.com and generate one on your dashboard.

Pass it as a header:

X-API-Key: fapi_your-api-key-here

Or as a query parameter:

curl "https://api.dev.thefeedapi.com/v1/news?symbol=BTC&api_key=fapi_..." | jq

Rate Limits

Every response includes rate limit headers so you can track your usage:

When exceeded, the API returns 429 Too Many Requests with a JSON body containing retry_after (seconds) and reset_at (ISO 8601). Free tier resets daily at midnight UTC.

List Articles

Returns paginated articles from all 32 sources, filtered by category or symbol.

curl "https://api.dev.thefeedapi.com/v1/news?symbol=AAPL&limit=2"

Click "Send Request" to try the API

List Sources

Returns all 32 RSS feeds being polled, with their URLs and category tags.

curl https://api.dev.thefeedapi.com/v1/sources | jq '.[].name'

Health Check

Simple endpoint to verify the API is running and get the total article count.

Python Example

Fetch news for a specific symbol and print matching articles:

# fetch_news.py
import requests

API = "https://dev.thefeedapi.com"

# Get all Bitcoin-related news
resp = requests.get(f"{API}/api/v1/news", params={
    "symbol": "BTC",
    "limit": 10,
})
data = resp.json()

# Print usage info from headers
print(f"Remaining: {resp.headers.get('X-RateLimit-Remaining')}")
print(f"Reset at:  {resp.headers.get('X-RateLimit-Reset')}")
print()

for article in data["articles"]:
    tickers = [s["ticker"] for s in article["symbols"]]
    sent = article["sentiment"]
    print(f"[{article['source']:20s}] {article['title'][:60]}")
    print(f"  Tickers: {tickers}  Sentiment: {sent:+.2f}")
    print()

Trading Bot Integration

Example of a simple contrarian strategy: bullish news → short, bearish news → long.

# Simple sentiment-based trading signal
import requests

resp = requests.get("https://api.dev.thefeedapi.com/v1/news", params={
    "symbol": "BTC",
    "limit": 5,
    "since": "2026-05-14T22:00:00Z"
})

avg_sentiment = sum(
    a["sentiment"] for a in resp.json()["articles"]
) / max(len(resp.json()["articles"]), 1)

if avg_sentiment > 0.5:
    print("Signal: SHORT BTC (market overly bullish)")
elif avg_sentiment < -0.5:
    print("Signal: LONG BTC (market overly bearish)")
else:
    print("Signal: NEUTRAL")

This pattern works with any symbol — AAPL, ETH, XAU, TSLA, SOL — and any timeframe.

Quality Scoring

Every article now includes a confidence score (0–1) and quality_signals object to help you gauge how much to trust each piece of news:

Example response snippet:

"confidence": 0.85,
"quality_signals": {
  "source_reliability": 95,
  "freshness_score": 0.99,
  "is_headline": true,
  "article_length": 224
}

Pagination

Use limit and offset to page through results. The response includes next_page when more results exist.

The next_page field preserves all existing query parameters (category, symbol, since).

Symbol Reference

The API currently indexes 280+ tickers across 5 asset classes. Here's a sample:

Search by ticker only (?symbol=BTC), not by the long name. Both ticker and full name are matched during ingestion.

Error Codes

Error responses are JSON with an error field.

Changelog

2026-05-15

• ✅ Quality scoring: confidence (0–1) and quality_signals on every article
• ✅ Source reliability scores (Bloomberg/FT = 95, WSJ = 93, NYT = 90, etc.)
• ✅ Freshness decay score per article
• ✅ Headline detection and article length signals

2026-05-14

• ✅ Symbol extraction engine launched (280+ tickers, 5 asset classes)
• ✅ ?symbol= query parameter for filtering articles by ticker
• ✅ Rate limiting: 1,000 requests/day per IP with X-RateLimit-* headers
• ✅ Pagination via limit + offset + next_page in responses
• ✅ 774 historical articles backfilled with symbol tags
• ✅ 32 RSS sources live and polling every 60s
• ✅ Sentiment scoring on every article

Earlier

• ✅ MVP deployed: Go + SQLite + Cloudflare + Let's Encrypt
• ✅ SSL/HTTPS with automatic HTTP→HTTPS redirect
• ✅ CORS enabled for browser-based access