FN2

FN2 is an AI stock research tool that runs autonomous agents against live market data on a schedule. Rather than opening a chat window and asking questions one at a time, you set up research loops that watch your portfolio, monitor earnings, and track macro conditions, then deliver a finished briefing before you sit down. The framing on the site is that your morning should start with a briefing, not a research session.

The site is unusually direct about the mechanics. An agent is a loop, a model that calls tools, reads the result, and decides the next step. FN2 runs that loop against live market data on a fixed schedule and grounds every step in real numbers instead of a model's memory. The transparency claim follows from that design, since you see every tool an agent calls and every source it pulls on every run, which the site sums up as no black box. Data sources are described as public and verifiable, every answer carries clickable citations, and every agent run is logged.

Five prebuilt agents ship ready to run, each a structured workflow rather than a chatbot prompt. Morning Briefing runs weekdays at 7:00 AM with a premarket read on your watchlist, overnight moves, the macro calendar, and the alerts that matter, using four tools in about eleven seconds. Earnings Monitor runs at market close and produces a beat and miss scorecard with CEO quotes, guidance changes, and signal tags across six tools. Thesis Tracker runs Fridays at 9:00 AM and follows supporting and contradicting evidence for an investment thesis, surfacing what changed that week. Macro Watchdog tracks rate expectations, yield curves, VIX, and macro releases, alerting on regime shifts that affect your holdings. Sector Scanner ranks a sector by fundamentals and flags value entries alongside deteriorating positions. You can also describe a custom agent in plain language and let FN2 build it, pick the tools, and run it on your schedule.

The data pipeline is the substance behind those loops. FN2 connects to Polygon.io for real-time prices, volume, and fundamentals, SEC EDGAR for 10-K, 10-Q, and earnings calendar data, FRED for 44 macro indicators updated daily, and Polymarket. It holds millions of earnings transcript segments from thousands of companies, adds AI-curated financial news in real time, and layers vector search for semantic retrieval across the documents. The argument it makes is concrete. When an agent reports a company's quarterly revenue, it pulled that figure from the actual earnings transcript rather than recalling it from training data, which is exactly where general-purpose chatbots tend to go wrong on financial questions.

There's an open API, which is the part developers will care about. You can point any agent at FN2 and pull grounded market research on demand, with live quotes, SEC filings, and earnings transcripts all sourced, and the site names Hermes Agent and OpenClaw as examples alongside anything you build yourself. It works in the other direction too. Dropping in the FN2 plugin lets your own agent create, schedule, and run FN2 loops on its own, so it manages a whole research desk instead of asking one question at a time. The division of labor the site describes is that your agents handle the workflow while FN2 handles the finance.

It's built for individual investors who want the kind of standing research coverage that normally sits behind institutional terminals. Founder Andreas Echavez says he built FN2 because he wanted a better tool to understand his own stocks, and because the best financial research tools were locked behind six-figure Bloomberg terminals and institutional trading desks while individual investors deserved better. Coverage centers on US public markets, which follows naturally from the sources it draws on. The site is explicit that this is not financial advice, that results are provided for educational purposes only, and that investment decisions are yours, so treat it as a research tool and do your own due diligence.

Pricing is freemium across three tiers plus enterprise. Free costs nothing, needs no credit card, and gives one active agent slot running daily with in-app delivery and full access to the data set. Pro lists at $29 a month, discounted to $7.99 for the first three months, and raises you to seven agent slots, one hourly and six daily, with in-app, email, and Discord delivery, scheduled earnings alerts and thesis tracking, advanced charts and exports, and priority research capacity. Max runs $89 a month for up to 21 slots, three hourly and eighteen daily, plus priority real-time alerts and priority support. Enterprise adds custom slots, dedicated delivery through Slack or webhooks, dedicated infrastructure, and volume pricing. Every plan includes the full data set, so the tiers gate frequency and delivery rather than access, and the site says you can have a first agent running in 30 seconds.