I didn't bring this up when it was a news story last month because there was a lot of cynicism in the thread, but I am genuinely curious. I am really grateful for both GrapheneOS and Google for creating a phone platform that Just Works for the essential stuff and that I can reasonably recommend to non-technical people!
[0]: >>44259921
GrapheneOS typically ports to new yearly Android releases in a couple days and tends to have it reach the Stable channel in under 2 weeks. We completed our initial port to Android 16 in a similar time period after the release on 2025-06-10. However, we then had to reimplement device support in a similar way to how we would support a non-Pixel device. Our initial production release based on Android 16 was published on June 30th. As usual, we had to spend around a week making a series of releases fixing regressions reported by users. It reached our Stable channel on July 8th.
Since our port to Android 16 took significantly longer than usual, we backported most of the Android 16 firmware, all of the kernel drivers and parts of the userspace device support to our now obsolete Android 15 QPR2 branch and did a few more releases based on Android 15 QPR2 where we were able to provide the full 2025-06-05 patch level which also turned out to be the full 2025-07-05 patch level due to no vulnerability fixes in the July 2025 Android Security Bulletin or Pixel Update Bulletin. This was an unusual approach and not generally a reasonable way of doing things. We were able to do it successfully.
It won't be nearly as much of an issue going forward since we dealt with building the new automation we needed. Our port to Android 16 QPR1, Android 16 QPR2, Android 16 QPR3, Android 17, etc. shouldn't be nearly as difficult and we should get back to our typical porting time for major releases.
Android and Chrome are potentially going to be split from Google:
https://www.nytimes.com/2024/11/20/technology/google-search-... (https://archive.ph/egRL4)
Pixels are no longer the Android reference devices. An Android company ending up with the OS, Google Play and Google's OEM partners wouldn't need Pixels. That's a possible reason for the change. However, the simplest explanation is that they're continuing to take cost cutting to an extreme where it negatively impacts their long term revenue far more than the money it saves. A lot of Pixels were sold due to first class support for using other operating systems including it not voiding the warranty.
Is there any chance that you fabulous guys could lobby for a smaller <5 inch phone with that OEM? (reference >>44586723 )
I bought a Pixel 9 Pro Xl specifically to use with GrapheneOS. Unfortunately, its OLED and my eyes were incompatible. The PWM on the screen was terrible and I had to return it after some headaches.
Of course, none of that was the fault of GrapheneOS. I absolutely loved using it and think your project is vital.
But I'm still left a bit confused about the future devices GraphaneOS will support:
Because you said discussion are being done with an OEM, will GraphaneOS switch from pixels to a different device?
You also said that not having the device tree won't be a major hurdle in building GraphaneOS for the future, does that mean we can expect the pixel 10 to have GraphaneOS or it's too early to know ?
Thanks again!
"AOSP needs a reference target that is flexible, configurable, and affordable — independent of any particular hardware, including those from Google." [0]
Emphasis on independent of any particular hardware.
Current speculation/inference suggests it is because of the antitrust case against them, preparing for the possibility that they may be divested of Android (or at least to decouple in meaningful ways [1]).
[0]: https://www.androidauthority.com/google-not-killing-aosp-356...
[1]: https://www.bloomberg.com/news/articles/2024-11-18/doj-will-...
Ahaha, don't get your hopes up, friend. The possibility of an adequate, degoogled Android with picky requirements as GOS on good, ultramobile hardware (matte DCI-P3 IPS, 3.5 mm audio, USB-C 3.2 or better, dedicated, ideally quick-access mSD card slot, IP68 rating, good cameras, EMR pen compatibility, changeable battery, non-plastic case) is virtually nil. That would essentially be a modern hybrid between a Samsung XCover Pro 6 and one of the older Samsung Note phones, e. g. the Note 9. Days long gone... :(
A collaboration with an OEM doesn't mean we'll stop providing existing or future Pixels if they continue to meet our requirements.
However one must note also that for now not even Snapdragon fulfills GOS requirements. If/when that changes, Snapdragon devices may have more open-source community momentum than Google's Tensor. Plus all the economy of scale, etc..
In terms of security, Microtek is even more far behind Snapdragon.
Again, not an Android dev here, take the text above with a grain of salt, YMMV, etc..
Pixels will be supported until end-of-life. Future Pixels will be supported if they continue meeting our requirements.
> You also said that not having the device tree won't be a major hurdle in building GraphaneOS for the future, does that mean we can expect the pixel 10 to have GraphaneOS or it's too early to know ?
We expect 10th gen Pixels to meet our requirements and we should be able to add support for them. It's not going to happen in 12 to 48 hours from the official launch of the devices as we did for around the Pixel 6a and later. It will be more work. We've automated most of the device support for existing Pixels now and have removed nearly all of the Android 15 QPR2 device trees rather than manually updating them. We're continuing to automate more and will use that approach for supporting new Pixels.
The devices with an OEM partner are further in the future than the Pixel 10. We need Qualcomm's new SoC with hardware memory tagging support to launch because a flagship Snapdragon is the best fit other than the current lack of hardware memory tagging. Some things need to be addressed by the OEM including licensing extra things like Qualcomm and filling in some missing features. There needs to be a clear, workable plan for updates including Linux kernel LTS branches.
I like being able to reach across the keyboard one-handed (to shift, etc) and I can't do that on modern, larger phones.
I've never owned a pixel, but I'm planning on getting the pixel 10 so I can use the new android 16 Linux terminal with graphical support. I hope the pixel 10 will still fit your requirements and it won't be a hurdle to port GOS without the device tree. I see no other OEM supporting the level of virtualization with kvm and vergil Google is bringing to the pixels. Good luck!