Innovative Eyewear has added Anthropic’s Claude to its Lucyd smart glasses, the company announced on July 10, 2026. Every Lucyd owner can now pick Claude or ChatGPT directly inside the companion app — at no additional charge. But the upgrade comes with a notable gap: you’ll need your phone in hand to start a Claude voice session, while ChatGPT already works with the phone locked and tucked away.

Two AI Assistants, One App

The centerpiece of the update is the Lucyd app for iOS and Android. It now surfaces a model selector that lets you move between Claude and ChatGPT without dropping the conversation’s context. That means you can ask one assistant to draft an email, then switch to the other for a fact-check without repeating the original request.

All Lucyd models — from original frames to the newest designs — get the integration. The glasses themselves are Bluetooth audio wearables, not head-up displays. They pipe voice responses into your ears and pick up spoken queries via built-in microphones. Visual output, image generation, and document handling remain on the phone screen.

The Hands-Free Gap

The catch sits squarely on how you start a Claude conversation. For now, you must pull out your phone, unlock it, open the Lucyd app, and manually choose Claude before you can talk. Once the session is active, the glasses handle the rest over Bluetooth.

Innovative Eyewear says a hands-free mode for Claude is coming in late Q3 2026. That update will let you invoke the assistant with the phone in a pocket or bag — matching the existing hands-free behavior that ChatGPT users already enjoy. The company hasn’t explained the technical hurdle, but it likely involves giving Claude the same background-voice-trigger permission that the OpenAI integration already has on mobile platforms.

Until that lands, Claude on Lucyd is best thought of as an app-driven feature that happens to route audio through your glasses. It’s not a fully off-screen, always-listening companion.

Features Under the Hood

Beyond the model switcher, the updated app bundles several capabilities that apply to both AI providers:

  • Image and PDF uploads: Hand a photo or document to the active model for analysis. The glasses don’t capture these directly; you select files from the phone’s gallery or storage.
  • Image generation: The prompt produces visuals that appear on the phone, not inside the lenses.
  • Saved chat histories: Conversations persist across sessions, though you can delete them manually.
  • Formatted code and tables: Useful for developers or anyone working with structured data on the go.
  • Incognito Mode: Sessions marked incognito are not retained after they end, at least according to the vendor’s stated behavior.

The app’s store listing also references upgrade prompts for “premium chat access,” so the free tier may impose daily message caps or limit some advanced features. Lucyd hasn’t published a detailed comparison table, but users should expect at least the basic conversational chops of each model without extra payment.

Who Benefits and Who Should Wait

If you already own Lucyd glasses, the update is a no-brainer. Update the app, check the model selector, and you instantly gain an alternative AI personality. Switching mid-conversation works reliably in testing, and having Claude’s more analytical tone alongside ChatGPT’s broader knowledge base can genuinely improve the experience of working through a complex task.

If you’re shopping for smart glasses, the delay on hands-free Claude changes the buying math. The promise of late Q3 2026 is a few months away, but there’s no guarantee it will arrive on schedule or work as smoothly as the ChatGPT implementation. Anyone who prioritizes truly hands-free operation over choice might still lean toward other wearables — or at least keep the phone within reach.

The Lucyd hardware is fundamentally an audio accessory; it doesn’t project a display. So the AI integration boosts its usefulness during walks, commutes, or while cooking, but it won’t replace a screen for reading long responses. That’s intentional, and it keeps the frames lightweight and affordable compared to full augmented-reality headsets.

Windows Users: What Changes (and What Doesn’t)

Lucyd glasses pair with a Windows PC as a standard Bluetooth headset. You can use them for Teams calls or music, but the AI features stay on the paired phone. There’s no desktop companion app or browser extension that routes Claude queries through the glasses while connected to a laptop.

That doesn’t mean the integration is irrelevant to PC work. Many professionals spend their day in front of a monitor while wearing Bluetooth glasses for call clarity. If you’re also using the Lucyd app to fire off quick AI queries during breaks, the glasses provide a convenient audio path. The phone simply remains the computational hub.

IT administrators should note two things. First, Incognito Mode only signals that the vendor won’t save session logs; it doesn’t encrypt files end-to-end or prevent the AI provider from processing data according to its own privacy terms. Employees should not feed confidential documents or proprietary images into a consumer AI service without approval. Second, the “free” label doesn’t necessarily mean unlimited. Check the app’s premium prompts and consider whether usage could hit a cap that interrupts workflow.

The Road to Multi-AI Wearables

Lucyd’s move is part of a broader pattern: accessory makers are racing to put multiple AI models onto devices that people already wear. Rather than betting on a single provider, they’re building a front end that can swap between engines based on user preference or task fit. That approach mirrors how desktop apps and web services have started offering model pickers, but it’s still rare in wearables.

For Anthropic, landing on Lucyd puts Claude into a physical product category where OpenAI already had a foothold. Winning real-world voice interactions could help Claude prove its conversational chops beyond text-based chat. And for Innovative Eyewear, supporting both models reduces the risk of being tied to one AI partner’s roadmap or pricing changes.

The lack of a hands-free launch for Claude suggests the integration is still maturing. Voice assistants that wake reliably from a locked phone require deep hooks into the mobile OS, and each provider’s SDK handles those differently. The late Q3 2026 target gives a rough sense of when those technical pieces might come together.

What to Do Right Now

If you have Lucyd glasses, open your phone’s app store and update the companion app. After the install, look for a model dropdown or button labeled “Claude” near the top of the chat interface. If you don’t see it, force-quit the app and restart; the rollout may be staged.

Once active, test the switching behavior. Start a conversation with ChatGPT, then mid-flow ask it to hand over to Claude. The app should retain the thread’s context. Experiment with image and PDF uploads to see how each model handles them — Claude may be better at dense document analysis, while ChatGPT might generate more polished code.

If Incognito Mode matters to you, toggle it on and verify that the session doesn’t appear in your history after you close it. But remember: this isn’t a substitute for proper data governance. Treat any upload as if it could be processed on external servers, because it will be.

For IT teams, the pragmatic step is to remind staff that Lucyd glasses are bring-your-own-device accessories. If the organization already permits voice assistants like Alexa or Google Assistant on phones, Claude on Lucyd doesn’t introduce a fundamentally new risk. The watchpoint is document sharing: make sure data classification policies cover AI-powered apps on personal devices.

What Comes Next

The hands-free Claude update, targeted for the end of September 2026, will be the real test. Once users can summon either assistant without touching their phone, the multiple-model promise becomes much more practical. Keep an eye on app release notes and the company’s social channels for a specific date.

Beyond that, Lucyd’s multi-AI approach could pressure competitors like Meta’s Ray-Bans or Amazon’s Echo Frames to offer similar model choices. And it raises a question for Microsoft: when will Copilot find its way onto third-party wearables, or does Redmond plan to keep its AI tightly coupled to Surface devices? For now, the answer is that wearable AI is becoming less about what one company can build and more about how well a device can juggle whatever models users trust.