Finterm
A command-line tool that feeds AI coding assistants and trading bots clean, structured financial data — earnings, SEC filings, insider trades — with a citation attached to every fact, so an AI agent analyzing a stock can show its sources instead of just m
🔗 Visit Finterm
Description
AI models are notoriously bad at financial facts on their own — they'll confidently state a stock price or earnings number that's wrong or outdated, because they weren't trained on real-time data. Finterm exists to fix that specific problem for AI agents: it's a command-line data layer that structures financial information — ticker data, earnings, SEC filings, ownership flows — and returns it with source attribution and timestamps, so an agent like Claude or ChatGPT can cite exactly where a number came from rather than hallucinating it.
It bundles several distinct research tools into one subscription: ticker data combining earnings, guidance, price reactions, options sentiment and technicals; SEC filing diffs that highlight what actually changed between two filings; a "deep research" mode that filters out AI-generated spam and low-quality content from web results; ownership-flow tracking of insider trades (Form 4) and institutional holdings (13F); options sentiment analysis (put/call ratios, implied volatility); and a 0-100 composite sentiment/risk score. It outputs agent-readable YAML/JSON with source attribution built in, installs via npm, and is operated by DXDT Labs Incorporated. Pricing is $150/month for the first 1,000 users during a preview period, rising to $200/month after, with a 3-day free trial; the company positions this against the $300-450/month it estimates the comparable standalone data products would cost separately.
💬 Our review
The short version: for a trading agent or a serious quant workflow that specifically needs source-cited, agent-readable financial data rather than a human dashboard, Finterm fills a real gap — but $150-200/month is a meaningful ongoing cost that only pays off if you're actually building something that consumes structured data programmatically.
Against Alpha Vantage, Polygon.io, or IEX Cloud, which are the standard financial-data APIs developers reach for, Finterm's differentiation is being purpose-built for AI agent consumption specifically — the citation-first, agent-readable YAML/JSON output and SEC filing-diff feature aren't things those general-purpose data APIs prioritize. That said, those alternatives are often cheaper for raw price/volume data alone if citation and filing-diff analysis aren't what you need; Finterm's premium is for the synthesis layer (deduplication, source-tiering, spam filtering) on top of raw data, not just data access itself. The company's own claim that comparable standalone products would run $300-450/month is a reasonable framing if true, making Finterm's bundling a real discount, but as a company that explicitly disclaims providing investment advice and is young enough that its data quality and uptime under real trading load aren't independently verified yet, it's worth using the 3-day trial to stress-test the actual data accuracy on tickers you know well before committing a monthly subscription to a live trading pipeline. <!-- ai-generated -->
💰 Pricing
📊 Global score
🤖 AI-enriched data
150$/mois (tarif de lancement, premiers 1000 utilisateurs), 200$/mois ensuite. Essai gratuit 3 jours.
Pros
Sortie citée et sourcée — conçue spécifiquement pour la consommation par agents IA
Comparaison de dépôts SEC (filing diffs) qui met en évidence ce qui a changé
Filtrage anti-spam/contenu généré par IA sur la recherche approfondie
Regroupe plusieurs outils de recherche financière habituellement vendus séparément
Cons
150-200$/mois — coût significatif pour un usage individuel
Entreprise jeune, fiabilité et disponibilité sous charge réelle non vérifiées indépendamment
Décline explicitement fournir un conseil en investissement
Alternatives comme Alpha Vantage moins chères si seul l'accès aux données brutes est nécessaire