Hi everyone, I've had the Beryl AX for over a year. The router works great, but I'm very disappointed with the frequency of updates. The latest version (4.8.1) was released in 2025-08. Why? I thought your update policy was better. I think it's right to release new products, but it's absolutely wrong to ignore those who already have a very good product that's still on the market. I don't know yet if I'll sell the router and switch to other suppliers who are more customer-focused. I hope the staff can respond, and I send you my best regards from Italy.
This particular model is very well catered for in the Beta versions of the firmware. One of the better supported models actually with updates released as recently as less than a month ago.
1 Like
- The Beryl AX is solidly in the pipeline for active firmware development and I think is the sweet spot for affordable but highly capable travel routers. Mine has the latest beta on the MTK SDK and it does everything I want, and tons I don’t want. I will be experimenting a bit when I have time with the opensource versions.
- I get that there might be some frustration that some products are not in the same active pipeline, like the original Beryl and the Marble. I also get the disappointment that some products aren’t considered powerful enough for the same features introduced for others, while some that aren’t powerful enough still are being dragged along by newer products (the Mudi V2 with the Mudi 7 being an example). But the Beryl AX is not in those categories.
- The GL-iNet GUI is enormously convenient and very effective for the things I’m interested in. But it has a lot of stuff I have zero interest in. I have no interest in tailscale or zerotier; or filesharing; parental control. And the features that are of interest in a fixed router are of less importance in a travel router, and vice versa. So having each product being feature equivalent is a false idol to begin with. For a time there was the ability to tailor the firmware through imagebuilder, but that doesn’t exist anymore. So updates are not necessarily of interest to me (other than security backports).
- Among significant wifi router manufacturers, GL-iNet stands out for its focus on being built on OpenWrt. So in many, perhaps not all, cases, its products have the possibility of being run on vanilla OpenWrt. I have a bunch of Asus Broadcom-based routers that are fully capable but mostly at a dead-end and using Linux kernels years out of date. I started with a Mango, then moved to a Beryl, then a Beryl AX. The Mango with 16Mib flash can’t keep up with the new firmware, but I have been able to load the most current OpenWrt with exactly the things I need and nothing I don’t, and it has been most satisfactory. If I’m traveling with a roll-aboard to an Airbnb for two weeks, I’ll take the BerylAX, but if I’m traveling with just a backpack the Mango is extremely weight/capability effective. (Looking forward to the Mango V2!)
As a complete non-sequiter, I’m in Portugal. So congratulations on having the freedom to choose that our US colleagues don’t have.
1 Like
Hi,
I also own the GL-MT3000. However, it’s still receiving updates. The newer version, v4.9.0 and v4.9.1 for OpenWrt 25 is in beta, so it will be released to the public soon. I’m currently on the beta versions to participate in reporting bug issues. The last update was actually on July 22, 2026 (v4.9.1, beta 3). You can go to the download center and switch to the Beta tab if you want to test out the update. (GL.iNet download center)
Its got 4.9 beta (which has lots of iterations and effectively is stable).
Its got op24
Its also got op25.
Im not sure what else you need?
If you actually crave a release tagged “stable”, what features does the current 4.8 tree lack that needs an update ?
MT3000 looks to be one of the best supported models.
2 Likes
Fiber internet is much cheaper in Portugal than in the US, so you're winning on something.
I'm looking forward to v4.10.
Hi, a few months ago I tested OpenWrt 24 version 4.8.3 (which I believe will remain Beta since version 4.9.0 was released, also Beta). The result was a disaster: there was too much network instability, and the Wi-Fi range was limited due to the open-source drivers. (This is specified here: GL.iNet download center ).
After a week, I reverted to the stable version 4.8.1 and everything was resolved. Unfortunately, I had to reconfigure everything because the LUCI restore didn't work (would you be so kind as to tell me why it didn't work?). I'm convinced that Beta versions are intended for Beta testers to use on test devices; otherwise, they wouldn't be Beta, so I don't think I'll waste any more time testing version 4.9.0 Beta and will wait for the stable version.
To answer your question:
- I need stability and security, since OpenWrt updates very frequently. I would have hoped that GL.INet would provide 2 or 3 updates per year. Thanks.
Your choice.
Beta iterations get better over time and final series betas are often just extended RC testing and perfectly stable, often the same code base.
Again, if 4.8.x is working perfectly for your needs, why do you NEED updates? What features do they have that you want that isnt in the stable version?
(Also, as in all software “stable” can also crash and be unstable in certain scenarios).
1 Like
Hi, you keep asking me what new feature I need, but I didn't ask for any new features in my post! OpenWrt version 21.02 has been EOL for a few years. Gl.iNet is committed to providing updates for products that are not EOL.
Updates are necessary for any IT device. I hope I've covered everything, and I don't think I'll reply to you again.
So what do you want ? You’re still not answering. What is in the current old stable setup that isnt satisfactory that needs changing to need a new version?
You seem to be arguing for updates just “because i want one” with no real need or reason for it.
What doesnt the current version do that you need?
There are several active development branches - more than most of the other product lines. It takes time to roll out new firmware, new kernel, new drivers, new features. You dont just randomly brand a release “stable” and push it for no reason just to satisfy a user who doesnt even seem to know what they want.