In Anthropic’s lab, 950 Claude agents screened more than 200,000 enzymes in 21 hours
This story was corrected after publication (Oct. 9, 2026). Details
Anthropic says roughly 950 Claude agents ran for 21 hours and used 210 million tokens to find a previously uncharacterized enzyme system in bacteriophage DNA. Claude searched more than 200,000 reverse transcriptases, identified 3,500 possible systems and narrowed them to the 20 most compelling candidates; researchers review those candidates and choose which ones to test in the lab. The system, named ART, pairs a reverse transcriptase gene with a neighboring protein of unknown function and evenly spaced DNA repeats resembling a CRISPR array; what it does is not yet known. The company puts it at weeks to months of work for an expert. The experiments are done by human scientists in the lab, not by Claude.
Sources
- Anthropic, “Claude discovers a novel enzyme system with CRISPR-like repeats”, (anthropic.com)
Corrections and updates
- The version first posted on Instagram said researchers picked 20 of the 3,500 possible systems. According to Anthropic, Claude agents narrowed them to those 20 candidates, and researchers review them to choose which to test in the lab; the text has been corrected.
About this story
This story was posted on Instagram by @jarrus.tech on Oct. 1, 2026.
Spotted an error in this story? [email protected] · Instagram
Short link: thejarrus.com/en/claude-enzyme-screening
This story in Turkish: Anthropic’in laboratuvarında 950 Claude ajanı 21 saatte 200 binden fazla enzimi taradı
On the same topic
Robot tests: even the best model completed only 19% of 84 robot tasks, and 63 tasks were solved by no model
In RobotWorld, a benchmark released October 7, the top model, GPT-6 Astra, completed only 16 of 84 simulated robot tasks; no model solved 63 of the tasks.
OpenAI’s image model posts a near-perfect score on text-dense images
On UltraText Bench, a dense-text test led by Westlake University, OpenAI’s GPT Image 2 ranked first of 24 configurations with 99.35 out of 100.
HackerRank’s AI interviewer opens to all customers after more than 500,000 job interviews
HackerRank made its AI interviewer Chakra generally available to customers on October 5, after a roughly six-month beta with more than 500,000 interviews.