HN Query Tool: Query 600GB of Hacker News, ArXiv and other datasets with Claude.

What it does

Scry is a query tool that lets users ask questions of large public datasets through ChatGPT or Claude. The tool connects AI assistants to indexed data from Hacker News, arXiv, Reddit, LessWrong, Stack Exchange, Wikipedia, and prediction markets. A user writes a natural language question, and Claude or ChatGPT formulates a SQL or vector search query to find the answer across these datasets. The tool also offers an Alerts feature that monitors for specific criteria and sends notifications when conditions are met.

Who it is for

The platform targets researchers, analysts, and curious individuals who need to search across public records and forums. Users can range from hobbyists asking one-off questions to developers building agents that query data programmatically. Scry is positioned as a research tool for people whose questions can be answered by public datasets rather than proprietary or private information.

Pricing

Free for non-commercial use when capacity is available. Commercial queries cost a fraction of a cent to a few cents per question. Custom pricing applies for commercial use, custom data sources, or datasets requiring reviewed access.

How it stands out

The integration approach is the primary differentiator. Rather than building a standalone search interface, Scry works as a plugin or connector within existing AI assistants—ChatGPT, Claude, and various developer tools like Claude Code and Cursor. Setup takes roughly two minutes without requiring code. The tool can handle complex, multi-condition queries that traditional search cannot easily express: finding every mention of a compound without references to a certain patent, grouped by domain, filtered by date. The indexed datasets are substantial—over 105 billion rows queryable across multiple sources.

What a founder should check

A competitor should investigate switching costs for users already relying on Scry's integrations within their daily AI workflows. The incumbent advantage of embedding within ChatGPT and Claude means new entrants would need to match that convenience or offer materially better results. Second, verify whether the data freshness and update frequency claimed actually matter for the target use case—if answers are acceptable with month-old Hacker News or arXiv data, this constraint may not be a moat. Third, examine the pricing model's sustainability: at a fraction of a cent per query, margins depend on scale and infrastructure efficiency. A rival should test whether users will tolerate switching to a new interface if pricing is comparable, or whether the integration simplicity creates genuine lock-in.

Thinking of building something like this?

Every launch here is a competitor to somebody's idea. If yours is close, check it against the market before you build: the Full Check names the rivals, the prices and the gaps.