Scry
Scry: programmable internet search w/ congestion pricing
Last verified:
What is Scry?
Scry is a programmatic research platform and MCP server that enables AI agents to run complex SQL queries over billions of internet documents. It transforms disparate data sources like social media, research papers, financial markets, and web crawls into a unified, relational database for deep analysis.
Scry pricing
Pricing model: Freemium
Free when servers have slack capacity; uses congestion-based pricing when busy, or $0.05/second for agents without an account via x402.
Scry pros
- Allows AI agents to perform structured SQL queries on internet data instead of relying on traditional web search
- Supports complex data extraction and analysis across massive datasets like Reddit, arXiv, Hacker News, and SEC filings
- Accessible via MCP endpoint and HTTP API, integrating seamlessly with agents like Claude, ChatGPT, and Cursor
Scry cons
- Requires technical knowledge (SQL) to use effectively, as it is designed for developers and AI agents rather than casual users
- Costs can fluctuate based on system load due to its congestion-based micro-auction pricing model
Frequently asked questions about Scry
What is Scry?
Scry is an authenticated, read-only query service over public corpora that allows agents to write bounded SQL queries across internet data.
How is Scry priced?
It is free when the system has slack, but switches to a congestion-based micro-auction price when busy, with agents without accounts prepaying $0.05 per second.
What sources does Scry cover?
It covers billions of documents including Reddit, Hacker News, Wikipedia, arXiv, SEC filings, GitHub archive events, and Polymarket.