Classiq, the Tel Aviv-based quantum computing software company, announced an AI agentic layer that converts natural language instructions into structured, executable quantum programs, according to The Quantum Insider. The company calls it a “first-generation, expert-level quantum agent.”
The distinction from conventional code assistants: the agent runs on top of Classiq’s model-based quantum software platform rather than generating free-form code. That means outputs are validated, compilable, optimized for real hardware constraints, and hardware-agnostic.
How It Works
Users describe high-level goals, algorithms, or domain-specific problems in plain language. The agent then generates quantum programs using Classiq’s synthesis and optimization engine, which ensures the output is structured and maintainable rather than one-off experimental code.
The agent covers the full lifecycle of quantum application development: translating domain problems into quantum models, designing scalable algorithms across abstraction layers, optimizing circuits under real-world constraints, and iterating within structured development workflows. Because it operates on Classiq’s curated algorithm libraries and validated modeling layer, the agent can reason about quantum systems at a higher level than general-purpose LLM tools, according to the company.
Target verticals include pharmaceuticals, finance, aerospace, automotive, and quantum error correction.
“AI in quantum computing has so far been limited to helping write code,” CEO and co-founder Nir Minerbi said in the announcement. “Classiq created the foundational modeling and abstraction layer for quantum software. It’s the only stack designed to be natively understood by large language models, allowing AI to develop expert-grade quantum applications that are not just theoretical, but fully compilable and ready to run on real hardware.”
Company Background
Classiq was founded in 2020 by Minerbi, Amir Naveh (CPO), and Dr. Yehuda Naveh (CTO), according to Calcalist. The company extended its Series C in November 2025 with funding from AMD Ventures, Qualcomm Ventures, and IonQ, according to SiliconANGLE. Previous investors include HPE, HSBC, Samsung, Intesa Sanpaolo, and NTT.
Vertical Agents Beyond Code and Text
The announcement signals a broader pattern in the agent market: specialized agents entering technical domains where generic LLMs lack the domain expertise to produce reliable outputs. Quantum computing, with its unique constraints around circuit optimization, hardware compatibility, and error correction, is a domain where a validated modeling layer underneath the LLM could prove more useful than prompt engineering alone. Whether similar domain-specific agent architectures emerge in other specialized fields, such as chip design, materials science, or biotech, will indicate how far the vertical agent trend extends.