Apple’s latest operating system beta has once again ignited speculation about an upcoming wearable product. Developer Sam Henri Gold uncovered system prompts in iOS 27 beta 2 that appear to reference a device codenamed B790. The discovered file, a system prompt JSON, includes language that explicitly describes handling input from “two images from cameras on either side of user’s head.” This strongly indicates that Apple is developing a product that places cameras on both sides of the wearer’s head — a form factor that aligns with the long-rumored AirPods Pro with embedded cameras.
While some initially speculated that the code might point to Apple’s rumored smart glasses, the codename B790 and the camera positioning make a stronger case for AirPods. The company has been quietly working on this project for years, with multiple reports from credible outlets suggesting that a new generation of AirPods Pro would incorporate small cameras. These cameras would not be used for photography or video in the traditional sense; rather, they would serve as sensors to feed visual data into Apple’s Visual Intelligence system.
Understanding the B790 codename
The codename B790 is notable because it differs from the B798 codename reported earlier by Bloomberg. This discrepancy could indicate that Apple is working on multiple internal prototypes or that the codename has changed as the product moved through development stages. In Apple’s internal naming conventions, codenames often evolve as components are revised or as the product’s scope shifts. The B prefix is traditionally used for audio products (for example, the original AirPods were codenamed B288), so B790 fits into that lineage.
The system prompt file discovered by Gold is part of a larger asset collection related to on-device AI planning. It describes how the device should handle user queries about the environment — for instance, asking Siri to identify a landmark or read a menu. With cameras on both AirPods, the device could capture a stereo view of the surroundings, allowing for depth perception and accurate spatial understanding. This capability would be essential for providing contextual answers without requiring the user to hold up an iPhone.
Visual Intelligence on your ears
Visual Intelligence is Apple’s umbrella term for AI-powered visual recognition features that first appeared on the iPhone 25 and later expanded to the iPad and Mac. The technology uses the device’s camera to analyze objects, text, and scenes in real time, then provides relevant information or actions. For example, pointing your iPhone at a plant can identify its species, or aiming at a poster can offer to add the event to your calendar. Bringing this capability to AirPods would be a natural extension, allowing hands-free queries.
The integration of cameras into AirPods presents several engineering challenges. The cameras need to be extremely small and low-power to fit inside the compact earbud housing. They would likely be low-resolution sensors optimized for object detection rather than capturing high-quality images. Apple has filed patents describing earbuds with ambient light sensors that could double as basic cameras, and the company’s expertise in miniaturizing components (such as the U1 chip and the H-series audio chips) suggests they can achieve this. The lenses would face outward, away from the user’s ear, capturing a wide field of view of what the user is looking at.
Delayed but not abandoned
According to earlier reports, Apple originally planned to launch the camera-equipped AirPods in 2026. However, development delays — particularly related to software integration and the optimization of Visual Intelligence for a wearable form factor — pushed the timeline back. Current expectations place the debut in the fall of 2027, alongside the 20th anniversary iPhone. This delay aligns with Apple’s pattern of holding products until they meet the company’s quality standards.
The software challenges are significant. On-device AI needs to process visual data in real time while maintaining the low latency required for real-world interactions. The H3 chip, anticipated in the next AirPods Pro, is expected to include a neural engine powerful enough to handle these tasks without draining the battery. Additionally, the connection between the AirPods and the iPhone (or other Apple devices) must be robust enough to relay visual data when needed, while also allowing standalone operation for simple queries.
Competitive landscape
Apple is not the first company to explore camera-equipped audio wearables. Meta’s Ray-Ban Stories smart glasses include cameras, but they are built into the frame rather than the earbuds. Samsung has experimented with smart glasses as well. The AirPods form factor offers a distinct advantage: most users already wear earbuds for music and calls, so adding cameras does not require a new behavioral change. The privacy implications, however, are considerable. Apple has emphasized on-device processing and user consent in recent years, and it is likely that the camera AirPods will include a physical indicator (such as an LED) when the cameras are active, similar to the MacBook’s camera privacy light.
Another potential use case is augmented reality audio. With cameras capturing the environment, the AirPods could adjust the spatial audio profile based on real-world objects — for example, making a voice sound like it is coming from a specific corner of the room. This would blend the line between audio and visual computing, creating a more immersive experience.
What the system prompt reveals
The JSON discovery by Sam Henri Gold includes specific instructions for how the device should behave in different scenarios. One excerpt mentions analyzing the scene to answer questions like “What is this building?” or “Can you read that sign?” The prompt also suggests that the device should be able to distinguish between objects in the foreground and background, using the stereo image pair. This level of detail aligns with Apple’s ongoing investment in spatial computing, as seen with the Vision Pro.
Interestingly, the code also includes fallback logic: if the visual data is insufficient, the device will ask the user to reposition their head or move closer. This implies that the cameras have a limited effective range, likely optimized for arm’s length or slightly beyond. Apple’s software team has been training the Visual Intelligence models on large datasets of everyday objects to improve accuracy, and the AirPods version would inherit these models.
Market impact and user adoption
The introduction of cameras in AirPods would mark a significant evolution in the product line. Currently, AirPods are primarily an audio accessory; adding visual sensing turns them into a computing platform. This could drive a major upgrade cycle, especially among users who already own the AirPods Pro 3. The ability to get visual information without reaching for a phone is a convenience that many would find compelling. Analysts predict that the new AirPods could command a premium price, potentially over $300, given the additional hardware and software capabilities.
Developers will also play a crucial role. Apple is expected to release APIs that allow third-party apps to tap into the AirPods camera data, enabling uses like real-time translation of menus, identification of products in stores, or navigation assistance for visually impaired users. The health and fitness potential is another avenue: cameras could track the user’s surroundings during outdoor runs to provide safety alerts or monitor form during exercises.
Despite the excitement, questions remain about battery life and heat dissipation. Adding continuous camera operation would significantly increase power draw. Apple may implement a duty cycle where the cameras activate only when the user triggers a query (via voice or a tap) rather than running constantly. The company’s recent work on low-power image signal processors could help mitigate this. Additionally, the charging case may need to be redesigned to accommodate larger batteries or cooling elements.
As development continues, more leaks and code discoveries are inevitable. The B790 codename will likely be replaced by a public-facing marketing name, perhaps simply “AirPods Pro with Visual Intelligence” or a new sub-brand like “AirPods Vision.” For now, the evidence from iOS 27 beta 2 provides the strongest indication yet that Apple is moving forward with this ambitious project. The combination of audio and visual AI in a wearable form factor could redefine what people expect from wireless earbuds.
Source:9to5Mac News