AskChem: Claim-Centered Infrastructure for Chemistry Literature Synthesis

AskChem is a chemistry literature infrastructure that extracts atomic, typed claims from papers and attaches provenance such as DOIs, quotes, and evidence locators.

It indexes 2.4 million claims from 147,000 papers and adds faceted browsing, an evidence graph, and an exploratory taxonomy.

The system is accessible via web, REST, SDK, and MCP, letting researchers and AI agents search and verify evidence at the claim level rather than the document level.

In benchmark tests, grounding a GPT-5.5 reader in AskChem improved citation performance and produced fully resolvable DOIs.