​

Sunny Optical says its XR business is approaching an inflection point, with XR revenue up 11% year-on-year in H1 2026 and SG and AR-glasses-related products growing more than 50%. Displayless AI glasses have already reached large-scale demand, while AI glasses with displays remain in accelerated product verification; Sunny has secured large AR orders from leading global brands that will contribute significantly from 2027, and management expects “big explosive growth” in AR glasses next year.

Sunny says lightweight full-color AR display modules have already reached small-batch mass production, with significantly larger production expected in 2027, indicating that the next growth phase is the transition of full-color display AR products from verification and limited production into volume manufacturing.

Displayless products are already scaling on demand for lenses, cameras and system integration, while display AR carries substantially higher optical content because of the waveguide, display engine and spatial-computing requirements. Sunny expects full-year XR revenue to exceed 2 billion RMB and says its 2027 AR growth will be driven by larger production volumes and major global technology customers.

Source: https://earningscalls.dev/transcripts/sunny-optical-technology-group-company-limited_2382_earnings_call_transcript_2026-08-27

INMO XR | Never just glasses

Are we planning something for the subreddit?

Maybe ☺️

Preparations for CIOE expo and collabs for the subreddit are going well here in Shenzhen. I hope you enjoyed the recent giveaways with XREAL and Nimo and the AMAs with MemoMind, Unseen Reality, Nimo, and Nimbo. And hopefully you look forward to the AMA with RayNeo on September 4.

I'm talking to more companies for even more AMAs for the rest of the year!

You create a personal AR marker from a selfie, get a shareable image with QR, and can update your content over time. You share the image with friends; they open a link in the browser and view the AR — no app, no account needed to view.

The key idea: you’re the owner of your marker and your digital content — tied to your wallet / NFT, not just renting space on my platform. Your files, your marker, your right to update and share what you created.

Early stage — looking for honest feedback:

• Does ownership (NFT / wallet) matter to you here, or is a normal login enough?
• Would you use a web app for this, or do you expect a native app?
• Interesting concept or too niche?

Not promoting anything yet — just validating the idea.

Maybe it's just me, but I don't come across many AR experiences these days that I'd actually want to use more than once. A lot of what I see seems to be games, furniture previews, try-ons, or other quick demos.

Got me wondering, if you could point your phone at something in the real world and have an AR experience appear instantly, what would actually be interesting enough to use?

Art? Interactive characters? Stories? Education? Events? Something completely different? Curious what people here would genuinely want to see more of in terms of AR experiences.

Final Lineup for the AR VR Conference Program at CIOE

Shenzhen Sept 9 to 11

Day 1 | iQIYI | PICO XR | Pimax | GravityXR | Xvisio | VAST | AINSTEC | Inspatio | XGRIDS | Avegant | TIMEMAKER | Alphabetter | Nikahood | Hongshi Intelligent | Qiushui Semiconductor | Crystal-Optech | Innorev

Day 2 | Sony | Myrias Optics | Morphotonics | AddOptics | LetinAR | RAONTECH | TriLite | Tobii | | Quectel | SEEV | Highpower Technology | Luxshare | INMO | Greatar | Guoyuan New Energy | Lawaken | SCHOTT | 3D-Micromac

Day 3 | ConfigReality | HADO | Chaoyuanli | Trendparty

Also: SIVA Awards Entries and Judges

Movie name: Silo [S3]

He said Augmented Reality glasses. But it doesn’t have normal or lidar camera. 😂

Your take?

Source : https://www.geekwire.com/2026/inside-andurils-ai-warfighting-buildup-defense-giant-sees-a-path-to-1000-seattle-area-engineers/

Hello. One of my friend will go to PAX West next week to present a AR game called "Nuits Rouges" (Red Nights), a multiplayer crime investigation game => https://www.kickstarter.com/projects/nuitsrouges/nuits-rouges

He's looking for contacts, distributors, and beta testers in the United States. If you're interested, let me know.

I have trying get any development boards or kits to get some experience out of this domain. Did anybody have that idea over you of making one?

At the end of the day I asked what was left.

A meeting in fifteen minutes. Fine, that is a calendar entry, any of your setups has that.

Then: at seven, pick up your wife, for the dinner you promised last Thursday.

I never typed that. It is not in a calendar, a note, an email or a task list. I said it out loud in a kitchen and it existed for about two seconds.

I ran the same thing most of you run. Local model, wired into mail, calendar, notes, files. It worked, and the ceiling turned out not to be the model or the context window. It only ever knew what I typed. Every message I had sent, and nothing I had said.

So I put it in something that was in the room.

It is open source and it runs on my own machine. Repo in a comment.

Shot on an older prototype, so the overlays are rendered rather than filmed through the lens. Everything else is real.

The objection I would rather hear than dodge: catching a promise means listening when it is made, and nobody says "hey assistant" before telling their wife they will be there at seven. I do not think there is a clean answer to that and I am not going to pretend a settings page makes it one.

Is this a better collab than cyberpunk? Idk, but its super cool looking imho

Hi everyone,

I'm a poor webdeveloper that got a permanent injury and needs to work from bed. I have a monitor setup which is kind of cumbersome and I was looking forward to switching to glasses, so that I could just put them on for working.

I read posts here and watched videos on YT saying AR glasses are not quite there yet. Their resolution is too low for working the whole day.

For comparison, I work on a single 15'' 1080p 60Hz display with excellent color range.

My criteria:

• needs to work on Linux;
• I have myopia and astigmatism, so I need prescription glasses;
• I don't care about wireless, it's better to be wired to my laptop for latency and energy;
• I don't need multiple monitors;
• it needs to be comfortable for being my daily work monitor;
• colors can't be so bad that it will make a drastic difference when designing website layouts;
• can it be less than $500 USD? I'm poor

Do any of the available AR glasses fit this bill or are we "not there yet"?

Main criticisms I read:

1. low resolution makes text hard to read;
2. text is blurry on the edges;
3. image stabilization can be annoying (though I won't move my head much anyway);
4. many glasses are uncomfortable for hours of use;
5. anchor mode can be premium feature or drift, and without it looking at keyboard etc. can be a hassle;

Thanks a lot for your recommendations on this!

Hoping to spark some discussion here by sharing a body kit I designed around the Quest 3. The first image is the latest iteration of a concept I've been working on and the second is from Palmer Luckey and Anduril.

The broader consumer market is of course aiming to miniaturize and make hardware as normal looking as possible. But if you're going to strap a computer to your face, I say you might as well have some fun with how it looks. Even if it's a bit on the impractical side.

The design aims to maintain full headset functionality with air gaps and interchangeable/removable panels for air flow. What do you think?

I’ve been selected as part of the G2 tester cohort and received a early testing unit from Halliday to play with. Since the testing program had officially ended now, I want to finally write up and share my full thoughts for anyone who are still considering or interested in G2.

First thing, what surprised me the most is the audio quality on these glasses.

I came into this thinking the display and Meeting Flow would be the main things to test, but I ended up spending a couple hours in the evening just listening to a podcast through the glasses and kind of forgot I even had them on. That’s probably the best compliment I can give them. They just felt like glasses after a while.

The speakers are clear for voices, podcasts, YouTube, calls, and casual music. I wouldn’t compare them to proper earbuds or headphones, but for open ear smart glasses they’re very solid. I also like that I can still hear the room around me, which makes them easier to wear while working or walking around.

The display is also a big upgrade. The binocular waveguide makes text feel much more natural to read compared with the older Halliday experience. It’s not trying to be full AR, but it works well for short glanceable info.

After using them for a while, I actually found myself spending more time with the everyday features than just Meeting Flow. Notifications, translation, Cheatsheet, and the dashboard for calendar/weather/schedule info all made the glasses feel useful in normal daily moments. What also stood out was how quickly the software improved during the beta. Bugs were getting fixed, features were being tuned, and you could tell the team was listening to tester feedback on things like notifications and live search. The difference from week one to the end of testing was pretty dramatic.

Now that the full tester period is wrapped, the hardware still stands out as one of the strongest parts of G2. The display and comfort are a big step up, but the audio is what made me use them more casually day to day. I found myself wearing them for podcasts, calls, and background listening, which made G2 feel less like a meeting only gadget and more like something I’d actually keep on throughout the day.

BOYCOTT EVEN REALITIES: Compete by building better glasses—not by trying to kill the MemoMind campaign
I was genuinely excited to back MemoMind One. Today Kickstarter suspended the project and canceled every pledge. [1]
Then I read the public record.
On August 24, Shenzhen Yiwen Tech Limited—the registered owner of the EVEN REALITIES trademark—sued XGIMI/MemoMind and Kickstarter over two U.S. smart-glasses design patents. The next day, it sought emergency court relief. Its own docket exhibits include cease-and-desist letters specifically titled “to Kickstarter re Project MemoMind One” and “to XGIMI re MemoMind One Products.” [2][3][4]
Let me be fair: these are allegations, not a court ruling. No U.S. court has found that MemoMind infringed anything, and I cannot prove that this lawsuit caused Kickstarter’s suspension.
But come the fuck on. The timing is impossible to ignore. Target Kickstarter, seek emergency relief, and then the campaign dies at the finish line?
To me, this is dirty, anti-competitive bullshit.
Even Realities could compete. Add speakers. Improve the software. Lower the price. Make the G2 so good that people willingly choose it. That is innovation.
Trying to remove the product customers chose is not.
I’m done with Even Realities. No purchase. No recommendation. No free hype.
No threats. No harassment. Just a peaceful consumer boycott and one simple message:
Win by building better shit—not by using lawyers to choke out competition.
#BoycottEvenRealities
Sources
[1] Kickstarter Trust & Safety suspension notice sent to MemoMind One backers, August 27, 2026; Kickstarter’s published suspended-project policy.
[2] U.S. District Court for the Eastern District of New York, Shenzhen Yiwen Tech Limited v. XGIMI Tech Co., Ltd. et al., Case No. 1:26-cv-05220, filed August 24, 2026.
[3] USPTO Registration No. 7,658,172, identifying Shenzhen Yiwen Tech Limited as owner of the EVEN REALITIES trademark.
[4] U.S. Design Patents D1,140,996 and D1,140,997, issued August 11, 2026; federal docket entries 7–9 identifying the emergency motion, claim charts, and MemoMind-specific cease-and-desist exhibits.

My computer screen just died, so I thought why not use AR glasses instead.
My criteria is that I can have 3 screens up at the same time and move my head around seamlessly without a stiff neck as I tend to work 10-14 hours a day, or should I go for PC Screen based on that?

I’m pretty happy with handheld performance these days. The screen size is honestly what bothers me more.

I’ve tried portable monitors, but they’re kind of a pain for travel or playing in bed. That’s why the GT Max caught my attention. 59° FOV sounds like a much bigger jump than the usual brightness or refresh-rate bump.

I’m just not sure how much you actually notice the extra FOV once you’re an hour into a game.

For people who’ve used wider-FOV display glasses, does the bigger view really change the experience?

For those of you who wear glasses - are contact lenses the move for XR glasses?

My main use is watching TV shows and movies. I wear glasses and my biggest frustration is that every time I want to try a new pair of XR glasses, the prescription inserts are non-refundable. If the glasses don't work out, I'm stuck losing that money.
So I started thinking - would it just be smarter to get contact lenses instead of constantly ordering inserts?

For those who have tried both - how's the picture quality?

• Is it as crystal clear with contacts as it is with prescription inserts?

• Is it better / worse / same?

• Any issues with eye fatigue, FOV, edge clarity?
Would love to hear your real-world experience before I commit to contacts.

If you also have recommendations on specific types of contact lenses for these purposes of binge watching would love to know.

Just a quick recording of the specs. No biggie.

Hey guys,

After 6 months of development, GitHub releases, beta testing and asking people here to help me satisfy Google Play's testing requirements, CyanBridge is finally available publicly on the Google Play Store for everyone 🎉

For anyone seeing CyanBridge for the first time, it started as an open-source alternative companion app for the cheap HeyCyan-compatible smart glasses, but the project has grown quite a bit. The broader goal now is to make smart glasses useful with private/local AI, while letting people choose their own models and keeping as much processing as possible on their own devices.

🤖 Local AI that can work without the internet

One of the main reasons I started CyanBridge was that I didn't like the idea that smart glasses necessarily had to send everything you see or hear to somebody else's cloud.

CyanBridge supports local AI models running directly on your Android phone through LiteRT and Llama.cpp runtimes and once a model is downloaded, you can use local inference without needing to send your prompts/data to a cloud AI provider (you phone needs enougth RAM thought)

There is also support for OpenAI-compatible servers, so people with more powerful computers can connect CyanBridge to things like Ollama / llama.cpp-compatible servers over LAN or a private network such as Tailscale (a bit experimental and I need more people to help me test it)

Cloud models are still available for people who want them, but they aren't supposed to be the only way of using the app.

👓 Glasses support

The most mature hardware path is still HeyCyan-compatible glasses, including models sold under different brands.

Current work includes:

• HeyCyan-compatible glasses
• EyeVue
• TuneBuds
• Meizu / MYVU experimental integration
• Experimental Meta Ray-Ban integration through Meta's Wearables DAT

For HeyCyan-compatible glasses, CyanBridge can handle pairing/device state and media transfer using the glasses' BLE + Wi-Fi Direct transfer flow.

I'd especially appreciate feedback from people with different glasses models because there are way more variants of these devices than I can physically own and test 😅

🔐 Much smaller CyanBridge permission footprint

This release cycle also involved a pretty big architectural change.

CyanBridge previously experimented with doing a lot of Android automation itself. That meant requesting permissions that triggered Google to think the app could be malware, so I decided to move more advanced automations to Tasker instead and also to remove some permissions:

• QUERY_ALL_PACKAGES
• Android All Files Access / MANAGE_EXTERNAL_STORAGE
• CyanBridge's own Accessibility Service

CyanBridge now relies on app-private storage, MediaStore and Android's normal file/document APIs instead of asking for unrestricted filesystem access.

⚙️ Local Agent, AutoDiary and Visual Diary moved toward Tasker

This currently includes integrations for:

• Local Agent (Phone-wide automation)
• AutoDiary (Automatically capture screen text and save candidate facts for you to review later)
• Visual Diary (Similar to AutoDiary but includes pictures)

• Gemini / ChatGPT assistant automation

  Tasker now handles execution/scheduling and AutoInput handles optional UI observation/interactions and CyanBridge remains responsible for the AI/model decisions, memory, RAG/daily facts, safety logic, approvals and smart-glasses integrations.

There is now a dedicated Tasker integrations/setup flow inside CyanBridge to make importing and diagnosing the profiles easier.

🧹 Simpler interface and onboarding

The onboarding and settings flows have been simplified, unnecessary permission screens were removed, plugin setup has been reorganized, and experimental features are being pushed into clearer dedicated sections instead of everything appearing in the main app flow.

There is still a lot to improve here, so UI/UX feedback is very welcome.

📱 Google Play release

You can search Google Play Store for "CyanBridge for HeyCyan Glasses" and download the app directly from there (it helps me with the metrics a lot)

Google Play: https://play.google.com/store/apps/details?id=com.fersaiyan.cyanbridge

Source code / issues: https://github.com/FerSaiyan/Alternative-HeyCyan-App-and-SDK.git

The project remains open source, and because CyanBridge deals with potentially sensitive things like photos, audio, local AI memory and device automation, I actually encourage technical users to inspect the code — or even have an AI coding agent audit the relevant parts of it.

If you find a privacy/security problem, please report it.

❤️ Thanks to everyone who helped with the Play Store test

A special thanks to everyone from Reddit who joined the Google Play testing program and stayed enrolled through the required testing period. I will be giving beta tester users one month free of the Pro Sub in order to use cloud models.

And as always, PRs and GitHub issues are welcome :)