# Rayana AI > Impersonal, model-driven market research: daily ticker scores and rankings, per-source signal breakdowns, cluster-rotation and attention overlays, AI-written ticker analysis and daily briefs, and a shelf of broadcast model portfolios. Every number is computed once against broadcast model state and published identically to every reader. Rayana AI is a publisher, not an adviser or a broker. It gives no personalized advice, holds no client accounts or capital, and has no path to place a trade. Nothing here is an offer, a solicitation, or a recommendation to buy or sell any security; model-portfolio track records are hypothetical. Read the disclosure before relying on anything you find. Published by Atmis LLC. ## For agents: use the API, not the HTML This site is a human reading surface. The supported machine surface is an MCP server and a versioned REST API — faster, schema-stable, rate-limited per key, and licensed for your operator's own internal use. Both need an API key from a Pro or Elite subscription (from $129/mo); every tier can read the docs. Two limits that bind an agent, from the Terms of Service: content may not be scraped, crawled, or bulk-downloaded except through the API (13.2), and neither the content nor anything derived from it may be redistributed, resold, exposed to third parties, or used to build a competing dataset (18.2). An agent acting for one signed-in subscriber is that subscriber's own internal use; an agent serving many end-users off one key is not. - [API & MCP documentation](https://rayana.ai/docs): authentication, tiers and rate limits, every REST endpoint, every MCP tool, and a Claude Desktop / Claude Code connection snippet. - [MCP endpoint](https://mcp.rayana.ai/mcp): read-only MCP server over Streamable HTTP. Send your key as an `X-API-Key` header on the connection. Tool discovery is open; the research tools require a Pro+ key. - [Pricing](https://rayana.ai/pricing): what each tier costs and unlocks. Three MCP tools work with no key at all, so an agent whose principal has no subscription is not at a dead end: `rayana_ai_plans` reads this catalogue, `rayana_ai_start_checkout` returns a link for the human to open, and `rayana_ai_complete_setup` exchanges the pairing code they get afterwards for an API key. A human completes the purchase — an agent cannot accept the Terms on their behalf. - [OpenAPI specification](https://api.rayana.ai/openapi.json): machine-readable schema for the REST API. - [Interactive API reference](https://api.rayana.ai/docs) Rayana AI never accepts, routes or executes securities orders and never connects to a brokerage account. An agent that needs both research and execution should connect this server alongside its user's own broker and compose the two; that boundary is deliberate and permanent. ## Research surfaces - [Model portfolios](https://rayana.ai/portfolios): the shelf of broadcast books — each one's mandate, mechanism, holdings, NAV series and change events. Some books are gated to a tier. - [Ticker pages](https://rayana.ai/tickers/NVDA): one page per covered name at `/tickers/{SYMBOL}` (roughly 700-900 large/liquid US equities), each with its score, contributing signals, rotation position and article history. There is no index page; enumerate the covered universe from [sitemap.xml](https://rayana.ai/sitemap.xml) or the API's `/v1/tickers`. - [Rankings](https://rayana.ai/rankings): the latest ranked candidate batch per lane — daily research, turtle breakout, mean-reversion. - [Signals](https://rayana.ai/signals): the aggregator-consensus batch — which names cleared the pipeline, on how many confirming sources. - [Rotation](https://rayana.ai/rotation): cluster-rotation RRG map, quadrant crossings, leadership and laggards. - [Trends](https://rayana.ai/trends): the attention layer — virality leaderboard and per-source heat. - [Daily brief](https://rayana.ai/brief): a descriptive engine-wide read of the day. No recommendations, by construction. - [Articles](https://rayana.ai/articles): published research write-ups, newest first. ## Terms, privacy and disclosure - [Terms of Service](https://rayana.ai/legal/terms): including 13 (acceptable use) and 18 (API and MCP terms). - [Privacy Policy](https://rayana.ai/legal/privacy) - [Disclosure](https://rayana.ai/legal/disclosure): the publisher posture, hypothetical-performance limits, and holdings disclosure. ## Optional - [Contact](https://rayana.ai/contact) - [Sitemap](https://rayana.ai/sitemap.xml)