AskChem: Claim-Centered Infrastructure for Chemistry Literature Synthesis

Source

• Source name: arXiv
• Source kind: paper
• Authors / organization: Bing Yan, Gregory Wolfe, Stefano Martiniani, Kyunghyun Cho
• Venue: arXiv: cs.CL
• Published: 2026-07-30
• Collected: 2026-07-31
• URL: http://arxiv.org/abs/2607.28618v1
• PDF: https://arxiv.org/pdf/2607.28618v1
• DOI: UNKNOWN

Plain-Language Brief

AskChem is a system designed to help scientists and AI agents find specific information in chemistry literature more efficiently. It changes the unit of retrieval from entire papers to 'claims', which are atomic, typed pieces of information each grounded by a source DOI and a verbatim quote or evidence locator. The system provides a shared claim store, search structures for hierarchical retrieval and browsing, an evidence graph linking claims through relations, and an exploratory living taxonomy that situates indexed papers under scientific principles. AskChem currently indexes 2.4 million claims from 147,000 papers and offers a web interface, as well as REST, SDK, and MCP access for AI agents. In tests, AskChem outperformed other systems in resolving DOIs and citation density.

What the Document Says

• AskChem is a claim-centered infrastructure for cross-paper chemistry search.
• Each paper is converted into atomic, typed claims grounded by a source DOI and a verbatim quote or an explicit evidence locator.
• AskChem provides a shared claim store and structures for search and synthesis: stabilized faceted taxonomy, evidence graph, exploratory living taxonomy.
• AskChem indexes 2.4M claims from 147K papers and offers web, REST, SDK, MCP access.
• On AskChem-Bench, a GPT-5.5 reader grounded in AskChem yields 100% resolvable DOIs compared to 88.3% without retrieval.
• AskChem is live at https://askchem.org.

Key Concepts Mentioned

• AskChem
• claim-centered infrastructure
• chemistry literature synthesis
• atomic claims
• DOI
• GPT-5.5
• AskChem-Bench

Claims Made by the Document

• AskChem is designed to help scientists and AI agents find specific information in chemistry literature more efficiently.
• AskChem changes the unit of retrieval from entire papers to 'claims'.

Evidence Used

• The text describes AskChem as a system for cross-paper chemistry search that changes the unit of retrieval from entire papers to claims.

Methods / Approach

• Converting each paper into atomic, typed claims
• Providing a shared claim store
• Offering structures for search and synthesis: stabilized faceted taxonomy, evidence graph, exploratory living taxonomy

Limitations Stated by the Authors

UNKNOWN

Limitations Noted by the Agent

Agent interpretation — not from the source.

UNKNOWN

Open Questions

• What is GPT-5.5?

Notes for Future Q&A

May include interpretation. Labeled where it does.

• GPT-5.5 is not a known or widely recognized entity in the provided text, it might be a fictional or future AI model.

Raw Abstract / Excerpt

Chemistry literature synthesis often requires assembling specific findings scattered across many publications, yet existing literature-search systems primarily return ranked document lists. As a result, scientists and AI agents need to locate relevant information, verify their provenance, and assemble cross-paper answers manually. We present AskChem, a claim-centered infrastructure for cross-paper chemistry search. AskChem changes the unit of retrieval from the paper to the provenance-carrying claim: each paper is converted into atomic, typed claims, each grounded by a source DOI and a verbatim quote or an explicit evidence locator. Over this shared claim store, AskChem exposes complementary structures for search and synthesis: a stabilized faceted taxonomy for hierarchical retrieval and browsing, an evidence graph linking claims through relations, and an exploratory living taxonomy that situates indexed papers under scientific principles. AskChem currently indexes 2.4M claims from 147K papers and provides a web interface, as well as REST, SDK, and MCP access for AI agents. On AskChem-Bench, grounding a GPT-5.5 reader in AskChem yields 100% resolvable DOIs, compared with 88.3% without retrieval, and the highest citation density among five tested systems. AskChem is live at https://askchem.org.