Companies

Diode Computers, Inc.

diode.computer

Diode uses AI to design and manufacture custom circuit boards for hardware companies from specification through production.

HQNew York City, New York, United States
Employees1-50
2 active roles
Jobs checked 9h ago
Developer ToolsAI ApplicationB2B SaaSPre-Seed / Seed

About

Diode Computers designs and manufactures custom printed circuit boards for hardware companies, including customers in robotics, medical devices, and aerospace. Its differentiation is an AI-native, code-based workflow with reusable validated modules, automated component and manufacturing selection, and human electrical-engineer review that compresses PCB projects from months to days.

Market

Diode competes in PCB electronic-design automation and electronics-manufacturing automation, positioning itself as an AI-native, end-to-end alternative for hardware companies rather than only a schematic or layout application. Its differentiation is the combination of code-defined schematics, reusable modules, LLM-assisted engineering, automated verification, supply-chain-aware BOM generation, and direct manufacturing fulfillment. Flux and Quilter also apply AI to PCB workflows, while Altium, Autodesk, KiCad, and Siemens represent established EDA alternatives that generally center on design software rather than Diode’s integrated design-to-fabrication service.

Target Customers

Diode targets hardware companies and engineering teams building products in robotics, aerospace, edge computing, IoT, wearables, and related embedded-electronics markets. Its offering spans small project teams through enterprise engineering organizations, with buyers likely to be hardware, electrical, and product-engineering leaders seeking faster custom-PCB development from specification through volume production.

At a Glance

Problem

Diode Computers addresses the slow, error-prone PCB development process that constrains hardware innovation. Traditional workflows rely on siloed, heavyweight EDA tools, manual file exports, fabrication email exchanges, and repeated board spins when components change or designs fail design-for-manufacturing checks. The result is a development cadence that can stretch to one board per quarter, with wasted engineering time, lost calendar days, expensive rework, and sometimes an entirely new board. Diode’s founders describe the practical failure mode as PCBs that do not work on the first try, requiring microscope-level rework or a respin.

The central use case is helping robotics, medical-device, aerospace, and other hardware companies iterate physical products faster, particularly when they need domestic manufacturing or must scale fleets and meet demanding launch windows. Diode is especially relevant where a board revision delays the broader product, such as a robotics controller, sensor system, power system, or high-speed compute board.

Product / Service

Diode provides an end-to-end, AI-assisted circuit-board design and manufacturing service. Customers can describe their requirements, while Diode’s models generate manufacturable designs using reusable, tested modules. Its open-source, Pythonic/Zener-based representation makes schematics understandable to both engineers and language models; a Rust PCB compiler verifies designs in milliseconds, and the workflow covers schematic generation, layout, routing, bill-of-materials matching, design-for-manufacturing checks, quoting, and fabrication handoff. Human electrical engineers review designs for safety, compliance, and manufacturability, and outputs remain compatible with established tools such as KiCad, Altium, and Cadence.

The delivery model is closer to “circuit boards as a service” than standalone design software: Diode charges for custom design and manufacturing, translates designs into factory requirements, uses manufacturing partners, and ships working boards to customers. The benefit is a shorter path from specification to physical hardware—projects described as months-long can be completed in days, with Diode reporting a board design every four days and a reduction in revision counts from five to one for some customers.

Market

Diode competes in AI-assisted PCB design, electronic-design automation, and outsourced electronics manufacturing. Its direct software competition includes established PCB platforms from Altium and Cadence, while newer AI-native competitors such as Quilter automate PCB placement and routing. Diode’s differentiation is its broader end-to-end model: it combines code-native design, AI generation and verification, component and supply-chain management, manufacturing-aware optimization, human review, and physical board delivery rather than stopping at a CAD layout.

The company appears to be commercially active rather than pre-revenue: it says it charges for custom board design and manufacturing, names Physical Intelligence and Saronic as customers, and reports work with Fortune 100 companies, frontier startups, and Tier-1 manufacturers. Diode also says one engineer shipped more than 100 unique designs in the prior year. It was founded in 2024, entered Y Combinator’s Summer 2024 batch, and raised an $11.4 million Series A in July 2025, bringing reported total funding to more than $14 million. Public materials do not disclose revenue or a broader customer-count figure, so the available traction is best characterized as early commercial adoption backed by substantial venture funding rather than proven scale.

Founders & Leadership

Davide AsnaghiFounder
Co-Founder and CEO
Lenny KhazanFounder
Co-Founder and CTO

Funding History

2024-07
Seed / acceleratorUndisclosed

Y Combinator

2025-07
Series A$11.4M

Andreessen Horowitz (a16z)

Recent News

2026-04-03product
Diode wants to help the US make hardware at the speed of ...

A Diode social video highlighted the company’s open-source tool for converting circuit-board design into code so AI can read, check, and generate new designs in hours rather than days.

2026-01-07
Everything is Computer

Andreessen Horowitz’s article identified Diode among companies illustrating what may be possible for PCBs, sensors, motors, batteries, compute, and power electronics.

2026-01
Diode Computers

Forbes’ 2026 30 Under 30 coverage listed Diode Computers in the Manufacturing & Industry category.

2025-12-26partnership
Claude AI: Revolutionizing Electrical Engineering Design ...

Independent coverage described Anthropic’s collaboration with Diode as a strategic effort to improve Claude’s electrical-engineering capabilities.

2025-12-15
Claude + Diode Zener. Faster, Better PCB Reference ...

Gend reported that Diode was developing a code-first PCB design stack and had raised funding to accelerate AI-assisted board design.

2025-12-13partnership
Teaching Claude To Design Circuit Boards That Ship

Diode detailed its partnership with Anthropic to use its open-source Zener language for real PCB design, allowing Claude to draft schematics, verify designs, and build reference modules. The effort produced roughly 250 public reference designs in two weeks.

2025-12-12partnership
Making Claude a better electrical engineer

Anthropic announced that it partnered with Diode Computers to improve Claude’s ability to auto-generate PCB reference designs in the Zener language.

2025-11funding
Announcing our Series A

Diode announced its Series A financing, backed by Andreessen Horowitz alongside Y Combinator. The company said the funding would help scale its compiler and onboard more fabrication partners.

Active Roles

2
New York, NY, US/Engineering/35d ago
New York, NY, US/Engineering/35d ago

Business Model

Diode charges customers for custom circuit-board design and manufacturing. Its published pricing shows small, medium, and large fabrication tiers priced at $10,000, $20,000, and $50,000 per unit respectively, while enterprise packages with bringup and priority support use contact-based pricing.

Products

AI-powered end-to-end custom PCB design and manufacturing platformProgrammatic PCB design stack consisting of the Zener language and Rust-based `pcb` compilerAI specification generation, reusable reference-module library, design review, simulation, and electrical verificationBOM compiler, instant quoting, DFM-aware routing, and manufacturing fulfillment through a fab networkOpen-source reference designs and custom boards, including USB-CAN hubs, flight controllers, Jetson carrier boards, and other embedded hardware

Customers

Anthropic (publicly disclosed partner; not explicitly identified as a customer)

Tech Stack

AI/LLMs, including Anthropic ClaudeZener, a Starlark-based domain-specific language for PCB schematicsRust-built open-source PCB compiler and command-line toolchainKiCad-compatible EDA with export support for Altium and CadenceSPICE simulation, netlist analysis, and automated electrical validationAI-assisted BOM, design-for-manufacturing, routing, quoting, and fab-network automation

Competitors

Flux
Quilter
Altium Develop
Autodesk Fusion
Autodesk EAGLE
KiCad EDA
Siemens PADS Professional