# Can Claude Fable 5.2 design my electronic product? > Anthropic and Diode used Claude Fable 5.2 to design three working devices. What AI can now do across circuits, PCB layout, firmware and enclosures, and what still needs engineering validation before production. Canonical page: https://www.denotec.co.uk/insights/can-claude-fable-5-2-design-my-electronic-product - Published: 14 Nov 2026 - Reading time: 11 min - Tags: Product development, Hardware, Prototyping, AI - Author: DENOTEC Artificial intelligence is changing how electronic products are designed and developed. Tasks that traditionally required specialist engineers, including circuit design, PCB layout, firmware development and mechanical modelling, are increasingly within the capabilities of advanced AI models. The latest example is Claude Fable 5.2, which has demonstrated the ability to design complete electronic devices from relatively simple product descriptions. For startups and businesses looking to develop new hardware, this raises an interesting question. Can AI now design an electronic product from an initial idea through to a working prototype? The short answer is yes, in some cases. However, there are important differences between generating an electronic design, producing a working prototype and developing a product that is ready for commercial manufacturing. ## What Is Claude Fable 5.2? Claude Fable 5.2 is an advanced AI model developed by Anthropic. Released in November 2026, it is designed for complex technical work, software engineering and tasks that require extended reasoning. Unlike earlier AI models that primarily generated text, code and explanations, Fable 5.2 has demonstrated much stronger capabilities in working with engineering tools and physical design problems. This includes the ability to: - Develop electronic circuit designs - Select electronic components - Generate schematics - Design and route PCBs - Create 3D mechanical models - Develop embedded firmware - Prepare bills of materials - Identify and revise design problems The important distinction is that Fable 5.2 can work across multiple stages of engineering rather than simply providing suggestions or generating isolated pieces of code. Anthropic recently demonstrated these capabilities through a collaboration with electronics design company Diode, using Fable 5.2 to develop three complete electronic devices. ## What Has Fable 5.2 Actually Designed? To demonstrate its hardware engineering capabilities, Anthropic worked with Diode to build three electronic devices using Claude Fable 5.2. The projects included a USB speakerphone, a home voice-assistant hub and a handheld virtual pet. {/* IMAGE: three Fable 5.2 devices */} ### 1. USB Speakerphone The first project was a compact speakerphone designed to connect to a computer over USB-C. The device required microphones, a speaker and Bluetooth connectivity, including a PCB antenna. Fable 5.2 was able to develop the electronic specifications, select components, design the circuit, arrange and route the PCB, model the enclosure and write the firmware. The resulting prototype was assembled by Diode and worked without requiring electrical modifications to the design. ### 2. Voice-Assistant Hub The second project was a home voice-assistant hub. This involved a more complex electronic design with multiple components and communication interfaces. The model needed to consider component placement, PCB routing, power delivery and the physical enclosure. It also had to account for practical manufacturing requirements rather than simply producing a circuit that looked correct in a schematic. ### 3. Handheld Virtual Pet The third project was a small handheld electronic device designed to function as a virtual pet. This required the integration of electronics, firmware and a physical enclosure into a compact product. As with the other projects, Fable 5.2 produced the design and directed the human engineers through assembly and testing. According to Anthropic, all three devices worked without requiring electrical modifications to their circuit designs. These demonstrations represent a significant development in AI-assisted hardware engineering. Previously, AI models were generally much more capable of writing firmware than reasoning about the physical layout and electrical constraints of a complete electronic product. *Source: Anthropic, [Claude can design electronic prototypes from start to finish](https://www.anthropic.com/research/claude-electronics-design).* ## Can AI Design a PCB? One of the most interesting developments with Fable 5.2 is its ability to reason about PCB placement and routing. PCB design is more complicated than connecting components together. An engineer needs to consider: - Component placement - Track widths and clearances - Power distribution - Signal integrity - Ground planes - Thermal management - Electromagnetic compatibility - Manufacturing constraints For example, a high-speed USB interface requires careful routing to maintain signal integrity. Power supply circuits need appropriate component placement and short current paths. Sensitive analogue signals may need to be kept away from noisy switching circuits. These are not simply drawing tasks. They require an understanding of how electrical circuits behave physically. Earlier AI models could generate schematic designs and PCB layouts, but frequently struggled with these practical constraints. Fable 5.2 has demonstrated a much stronger ability to reason about component positioning and routing priorities. In Anthropic's testing, it also showed an ability to recognise when a design was becoming difficult to route and revise its component placement. This is particularly important because PCB design is an iterative process. A design that cannot be routed correctly may require changes to the component arrangement or even the circuit itself. AI models are becoming increasingly capable of handling this process. ## Can Fable 5.2 Develop the Firmware Too? Yes. Firmware development is another area where AI can contribute significantly. Embedded firmware controls how electronic hardware operates. It allows microcontrollers to communicate with sensors, manage power, process data and control external devices. Fable 5.2 can generate firmware for microcontrollers and work with the supporting software required to operate an electronic product. Depending on the hardware platform, this could include: - Initialising microcontrollers and peripherals - Developing I2C, SPI and UART drivers - Integrating sensors - Implementing Bluetooth and Wi-Fi communication - Managing power consumption - Processing sensor data - Developing USB interfaces - Implementing application logic - Supporting firmware updates It can also help identify problems in existing firmware and make changes based on testing results. However, generating firmware is not the same as proving that it works correctly on physical hardware. Microcontroller configuration, timing requirements, peripheral behaviour and hardware-specific limitations can introduce problems that are difficult to identify without actually running and testing the software. The most effective approach is to use AI to accelerate development while validating the resulting firmware against the real electronic design. ## Can It Design the Mechanical Enclosure? Electronic products rarely consist of a PCB alone. They often require an enclosure, mounting points, connectors, buttons, displays and other mechanical features. Fable 5.2 has demonstrated the ability to create 3D models of physical components and product enclosures. This is particularly useful for early-stage product development. For example, an engineer can provide the dimensions of a PCB and ask the model to develop an enclosure around it. The resulting design can account for: - PCB mounting positions - Connector locations - Button and switch placement - Display openings - Sensor positioning - Battery compartments - Overall product dimensions The designs can then be used to produce 3D-printed prototypes. There are still practical considerations around material selection, mechanical strength, tolerances, heat dissipation and manufacturing methods. A 3D model that looks correct does not necessarily mean the enclosure will be suitable for injection moulding or long-term use. Nevertheless, being able to develop electronics and mechanical components together is a significant advantage. ## What AI Still Cannot Guarantee Despite these advances, there is an important distinction between designing a working prototype and developing a commercially viable electronic product. The demonstrations from Anthropic are impressive, but they should not be interpreted as evidence that AI can independently take any product from an idea to mass production. There are several areas where engineering verification remains essential. ### 1. Electrical Validation A schematic or PCB layout can appear correct while still containing electrical problems. Component ratings, power supply requirements, signal levels and protection circuits all need to be checked. For example, a voltage regulator may be correctly connected but operate outside its thermal limits under the expected load. Similarly, a circuit may function under ideal conditions but fail when exposed to electrical noise or variations in supply voltage. These issues require proper engineering analysis and physical testing. ### 2. Component Availability and Selection Selecting the correct components is a major part of electronics development. AI can search for components and compare specifications, but component selection involves more than finding a part with the right headline features. Engineers need to consider: - Operating temperature - Electrical tolerances - Package dimensions - Supply voltage - Lifecycle status - Manufacturer support - Availability - Cost at production volumes Choosing an obsolete or unsuitable component can create significant problems later in development. Component datasheets and manufacturer documentation must be treated as the authoritative reference. ### 3. PCB Manufacturing and Assembly A PCB needs to be manufacturable, not just electrically functional. Manufacturing constraints include minimum track widths, spacing, drill sizes, copper thickness and board stack-up. Assembly introduces further considerations around component footprints, soldering requirements, connector accessibility and inspection. A board that passes basic design checks may still be difficult or expensive to manufacture. ### 4. Testing and Reliability A prototype working once does not demonstrate that it will operate reliably over several months or years. Commercial products may need extensive testing, including: - Electrical testing - Functional testing - Temperature testing - Vibration testing - Long-duration operation - Battery performance testing - Communications testing - Environmental testing The appropriate testing depends on the product and its intended application. Anthropic's demonstration projects were working prototypes, but their long-term reliability and commercial certification requirements had not yet been established. ### 5. Regulatory Compliance Electronic products sold commercially often need to meet regulatory requirements. Depending on the product and target market, this may include EMC testing, electrical safety, radio equipment requirements and other applicable standards. For example, products containing Bluetooth or Wi-Fi transmitters may require additional consideration around radio compliance. These requirements cannot be satisfied simply by generating a PCB design or firmware. They need to be considered during development and verified through the appropriate testing and assessment processes. ## Does This Mean You No Longer Need an Electronics Engineer? Not quite. AI is becoming a powerful engineering tool, but it does not eliminate the need for electronics engineering expertise. In fact, as AI becomes more capable of generating complete designs, the role of an engineer increasingly involves reviewing, validating and improving the output. An experienced electronics engineer can identify problems that may not be obvious from the generated design. They can also make decisions based on practical experience with component behaviour, manufacturing processes, debugging and product development. For example, an AI model might select a particular power supply circuit based on the electrical requirements. An engineer may recognise that the circuit will create thermal problems inside the intended enclosure or introduce noise into a sensitive analogue measurement. Similarly, a PCB might pass design rule checks but still have layout issues that affect real-world performance. The combination of AI-assisted development and experienced engineering judgement is likely to be much more effective than either approach in isolation. ## How Could AI Change the Cost of Electronics Development? One of the biggest potential benefits of AI in hardware engineering is the reduction in time required to produce an initial design. Traditionally, developing a custom electronic product involves several separate engineering activities. These might include: 1. Requirements gathering and system architecture 2. Circuit design and component selection 3. Schematic capture 4. PCB layout 5. Firmware development 6. Mechanical design 7. Prototype assembly 8. Testing and debugging Each stage requires specialist knowledge and engineering time. If an AI model can complete parts of these activities more quickly, the overall development process could become significantly more efficient. This is particularly relevant for startups and smaller businesses that may have limited budgets for initial product development. However, faster design generation does not necessarily mean a proportionate reduction in total project cost. Prototype manufacturing, component purchases, testing, certification and production preparation still involve real-world costs. There may also be additional engineering work required to review and refine AI-generated designs. The greatest savings are likely to come from reducing repetitive design work and accelerating early development rather than removing engineering from the process entirely. ## Should You Use AI to Build Your Next Hardware Product? AI-assisted hardware development is already becoming a practical option for certain types of electronic products. It is particularly promising for: - IoT devices - Sensor-based products - Connected consumer electronics - Microcontroller-based control systems - Simple monitoring equipment - Electronic prototypes - Early-stage product development For businesses developing these types of products, AI can help accelerate the initial design process and explore different technical approaches. It may also make it easier to test ideas that would previously have required a significant investment in engineering time. However, more complex products involving high-speed electronics, advanced power systems, safety-critical functionality or demanding operating environments still require substantial specialist engineering. The complexity of the product matters more than the capabilities of the AI model alone. ## The Future of AI in Electronics Development The ability of AI models to design complete electronic devices represents an important change in product development. For the first time, advanced models are demonstrating the ability to work across circuit design, PCB layout, firmware and mechanical engineering as part of a single development process. This does not mean that electronic product development has become fully automated. A generated design still needs to be manufactured, tested and validated before it can become a reliable commercial product. However, the time and resources required to move from an initial idea to a working prototype could become considerably smaller. For startups, researchers and established businesses, this creates new opportunities to experiment with hardware, validate product concepts and develop electronic products more efficiently. The most interesting development is not simply that AI can now design a PCB. It is that AI is beginning to understand how the different parts of an electronic product fit together. As these capabilities continue to improve, AI-assisted engineering is likely to become an increasingly important part of how electronic products are designed and developed.