Designing more efficient quantum circuits with Claude
Anthropic says Claude autonomously redesigned quantum circuits, beating published best designs by an order of magnitude in some…
Anthropic reports that Claude, given plain-language specs for eight quantum circuit design tasks and an automated correctness/efficiency evaluator, worked in a closed loop without human intervention to discover circuit designs using fewer qubits and gates than the best previously published algorithms. Highlighted results include a new modular inversion approach for elliptic-curve arithmetic and a roughly 12x cheaper version of the costliest step in a leading quantum optimization algorithm.
Why it matters
If reproducible, this demonstrates AI models can contribute novel algorithmic insight to a highly specialized, mathematically constrained domain (quantum circuit design) rather than just optimizing familiar classical code, which could accelerate the timeline for practical quantum algorithms by reducing the hardware resources they require.
What is verified
Anthropic gave Claude plain-language specifications for eight quantum circuits (from chemistry, optimization, physics simulation, and number theory) plus an automated evaluator, and Claude found designs that improved on the best published circuits, in some cases by an order of magnitude or more. [1]
Claude worked in a closed loop: proposing a circuit design, receiving a score from an evaluator program measuring correctness, qubit count and gate count, and using that feedback to improve, with no human input between rounds. [1]
In one example, Claude found a new way to perform modular inversion (the most expensive step in elliptic-curve calculations) and made the costliest step of a leading quantum optimization algorithm about 12x cheaper. [1]
The work was led by Andi Gu as part of Anthropic's STEM Fellows Program, with mentorship from Anthropic quantum physicist Alek Kemeny, in collaboration with researchers from Harvard University and QuEra Computing. [1]
The model used was referred to as Claude Mythos 5. [1]
What is not
- No independent replication or third-party review of the claimed order-of-magnitude improvements is provided; this is a vendor-authored blog post referencing an arXiv paper not itself examined here.
- The specific baseline comparisons (which published circuits, under what conditions) are not detailed in the text beyond the FeMoco and point-addition/modular-inversion anecdotes.
- No information on how these circuit designs translate to real near-term quantum hardware runs or whether they have been experimentally validated.
- The name 'Claude Mythos 5' is unusual and not otherwise corroborated as an official Anthropic model name in this evidence.
Sources
- 1Anthropic Newsanthropic.com · Primary source
Original source published 16 October 2026 at 19:29.
An AI-assisted briefing by Coding Partners. We link the evidence behind the reporting so you can explore the original work and assess the claims. How CP Signal works.