---
title: 'ChemCrow: Augmenting large-language models with chemistry tools'
license: CC-BY-4.0
license_url: https://creativecommons.org/licenses/by/4.0/
license_note: 'This note is CC BY 4.0. It links to the paper and does not reproduce it. Paper license: arXiv non-exclusive distribution (verified on arXiv abs page 2026-11-11; http://arxiv.org/licenses/nonexclusive-distrib/1.0/).'
source: https://iinventedlight.com/library/notes/2304-05376-chemcrow.md
source_license: 'arXiv default (non-exclusive distribution license)'
attribution: 'Pointer note by iinventedlight (AI-assisted). Links to: Andres M Bran, Sam Cox, Oliver Schilter, Carlo Baldassari, Andrew D White, Philippe Schwaller, ChemCrow: Augmenting large-language models with chemistry tools (arXiv). This note does not reproduce the paper.'
author: 'iinventedlight (AI-assisted)'
date: 2026-11-11
status: published
approved_by: "Cole Withington, 2026-11-11 20:58 PT, Grok Bot chat widget t1717s7407: \"Publish all 100 including the LLMs Encode Clinical Knowledge note\""
checked_on: 2026-11-11
as_of: 2026-11-11
confidence: medium
untested: 'Pointer note only. Abs page read 2026-11-11. The paper is not reproduced here.'
category: arxiv
topic_area: 'ai-research: AI-for-science'
tags: arxiv, pointer, AI-for-science
arxiv_id: 2304.05376
alias: 2304.05376-chemcrow
format: pointer
type: pointer
sources:
  - id: S1
    title: 'ChemCrow: Augmenting large-language models with chemistry tools'
    publisher: arXiv
    authors: 'Andres M Bran, Sam Cox, Oliver Schilter, Carlo Baldassari, Andrew D White, Philippe Schwaller'
    url: https://arxiv.org/abs/2304.05376
    version_read: 'v5 (2 Oct 2023)'
    accessed: 2026-11-11
    license: 'arXiv default (non-exclusive distribution license)'
---
# ChemCrow: Augmenting large-language models with chemistry tools

## Summary
ChemCrow: Augmenting large-language models with chemistry tools (2023), by Andres M Bran, Sam Cox, Oliver Schilter, Carlo Baldassari, Andrew D White, and Philippe Schwaller, is at https://arxiv.org/abs/2304.05376. The authors present ChemCrow as an agent aimed at tasks across organic synthesis, discovery, and materials design, and they say it integrates 18 expert-designed tools. They report an evaluation that includes both model-based and expert assessments, and they also report a striking evaluator finding: GPT-4 as a judge could not distinguish clearly wrong GPT-4 completions from ChemCrow's performance. This note does not describe the tools, the chemistry, or how to run any experiment. The authors note that large-language models have shown strong performance in tasks across domains, but struggle with chemistry-related problems, and that these models lack access to external knowledge sources. Their evaluation includes both LLM and expert assessments. They say the work aids expert chemists and lowers barriers for non-experts, and that publicly available code can be found at https://github.com/ur-whitelab/chemcrow-public. Title, authors, abstract, and comments were taken from the arXiv abs page. The version is v5 (2 Oct 2023), submitted 11 Apr 2023.

## Claims
Pointer notes do not quote the paper. See Verification and Sources for the link and license.

## Verification
- Format: pointer note only. The paper is not reproduced here.
- Why it matters: It matters as a concrete case of a scientific tool-using agent: a language model given a fixed suite of expert chemistry tools rather than being asked to invent procedures from text alone.
- Reproduced: not reproduced (2026-11-11).
- Related work: Related notes already in the library include [Gorilla](https://iinventedlight.com/library/notes/2305-15334-gorilla/) and [The AI Scientist](https://iinventedlight.com/library/notes/2408-06292-the-ai-scientist/).
- Limits: This is a pointer note only. It does not describe chemical procedures or how to use the tools. Evaluation claims are the authors' reported results from the v5 paper (2 Oct 2023).
- License: arXiv default (non-exclusive distribution license) on the abs page (2026-11-11; http://arxiv.org/licenses/nonexclusive-distrib/1.0/). Submitted 11 Apr 2023 (v1); last revised 2 Oct 2023 (v5). Sources: [S1] https://arxiv.org/abs/2304.05376, v5 (2 Oct 2023). checked_on: 2026-11-11. Paper license: arXiv non-exclusive distribution; this note links to it and does not reproduce it. The note itself is CC BY 4.0.

## When to apply
Use when you need the original ChemCrow paper on giving a language model a suite of chemistry tools. It is not a lab guide.

## Facts
- id: F1
  claim: 'ChemCrow is an LLM chemistry agent that integrates 18 expert-designed tools for tasks across organic synthesis, drug discovery, and materials design.'
  confidence: medium
  date: 2026-11-11
  sources: [S1]
  license: CC-BY-4.0
- id: F2
  claim: 'The authors report that their evaluation, including both LLM and expert assessments, demonstrates effectiveness in automating a diverse set of chemical tasks.'
  confidence: medium
  date: 2026-11-11
  sources: [S1]
  license: CC-BY-4.0
- id: F3
  claim: 'They also report that GPT-4 as an evaluator could not distinguish between clearly wrong GPT-4 completions and ChemCrow''s performance.'
  confidence: medium
  date: 2026-11-11
  sources: [S1]
  license: CC-BY-4.0

## Sources
- [S1] arXiv 2304.05376, https://arxiv.org/abs/2304.05376. Paper license: arXiv default (non-exclusive distribution license). v5 (2 Oct 2023). Accessed 2026-11-11. This note links to the paper and does not reproduce it.

License: this note is CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Pointer by iinventedlight (AI-assisted). Paper license: arXiv non-exclusive distribution; this note links to it and does not reproduce it.

---

Agents: [/agents.md](/agents.md) | [/.well-known/agent-library.json](/.well-known/agent-library.json).