alphai
REST API and MCP server for relevance-scored, ticker-linked financial news for trading bots and AI agents.
Last verified:
What is alphai?
alphai is a stock news API that delivers financial news articles scored by relevance (1–10) and linked to ticker symbols, designed for trading bots and agent backends. It serves pre-filtered content via REST API, MCP server, and official Python/TypeScript SDKs with server-side filtering by symbol, category, and relevance threshold.
alphai pricing
Pricing model: Freemium
Free tier (20 req/min, 100/day, no credit card required); Basic and Pro plans available
alphai pros
- Pre-filtered and relevance-scored articles (1–10) reduce noise—only ~20% of collected articles reach the API
- Server-side filtering by symbol, category, and relevance cuts bandwidth costs before the request hits your system
- Real-time ingestion: articles within ~1 hour of publication, SEC filings in ~6 minutes median
- Per-ticker sentiment analysis with confidence levels and reasoning included in every row
- Multiple access patterns: REST API, MCP server for agents, typed Python/TypeScript SDKs, or plain JSON with any HTTP client
alphai cons
- Free tier capped at 20 requests/minute and 100/day—suitable only for evaluation and prototypes
- Relevance scoring relies on LLM evaluation, introducing potential subjectivity despite validation checks
- Aggressive filtering means 80%+ of raw articles are dropped before API exposure, limiting comprehensive coverage for users who want full firehose access
Frequently asked questions about alphai
How current is the data?
Articles arrive within an hour of publication; SEC Form 4 and 8-K filings median ~6 minutes from filing.
Do I have to use the SDK?
No. SDKs are optional—you can send plain HTTPS requests with any HTTP client. SDKs handle cursor pagination, retries, and typed errors if you prefer them.
How are articles filtered and scored?
Each article is LLM-scored 1–10 for relevance, validated for ticker accuracy, and must clear the relevance threshold with at least one verified ticker to reach the API.
What data do I get per article?
Relevance score, category, extracted and verified tickers, per-ticker sentiment with confidence and reasoning, publisher URL and domain, and both publication and ingest timestamps.