FW 4.5.6 release 1 is in beta phase, you must download it and upgrade through web UI manually. (it’s quite stable in my short experience)
BTW I only have a decent wifi coverage when using firmware 4.5.2 or 4.5.3 beta, where the wifi transmission power is 36dbm
I don’t think the real power is really 36dBm, that’s a a lot of power, remember that dBm are logarithmic, if you convert them to mW you’ll have a better idea. I expect the true transmission power of the Flint 2 to be 30dBm (around 1000mW)
@alzhao Do you have information regarding the Flint 2’s transmission power?
You are right, it’s not really 36 dbm, I think it’s 30 dbm. The problem is now in the recent firmware, on luci shows 30 dbm but per my measure (using a spectrum analyser) it seems just 24 dbm. That’s why if you compare Flint 2 with other routers with the same country configuration, the wifi range of Flint 2 is shorter.
I wouldn’t be so sure, since I switched over to OpenWrt snapshots and noticed the dip in power when compared to 4.5.2. Then when GL.iNet released 4.5.3 I quickly tested that and found that my PCs signal strength was identical to what I get with OpenWrt snapshots.
What I’m really curious about though is if the routers firmware had the boosted transmit power “bug” during certification. Because if it did then it passed while the transmit power was boosted 
As a telecommunications engineer I worked for a few years in testing and measurement companies, carrying out these tests at regulatory agencies in my country (analogous to the FCC in the USA). In general, these tests are very demanding, it is difficult to know if there was such a “bug”
Yep. That’s what I was using right before I switched over to OpenWrt snapshots.
When switching to and from the stock firmware I’ve used inSSIDer to compare the signal strengths.
Yeah. It’s not something that anyone here can answer, but if the transmit power wasn’t boosted during the certification process then why was it changed after it’d passed certification? It’s very strange.
But in general I think GL.iNet should make their web UI and app ask the user what country they’re in when they first configure the router, since for many people the WiFi will be misconfigured by default and they’ll never access LuCI to correct the settings.
5 Likes
In using the latest 4.5.6 build 2024-01-18, after a few hours I lose DNS connection. If I reboot the router, or even disconnect / reconnect wireless at my PC I get he DNS connection back. Went back to the 4.5.4 release 2023-12-12 which is stable for me.
This is the first group where people complain about too much transmit power and the default country setting. Let’s keep this simple please. From LUCI, my default country is correct. However, if someone bought the router in a different country than their own, the default will be wrong of course. Same with any brand of router.
2 Likes
We nedd a more straightforward way of tracking the many bugs seriously, this it’s getting ridiculous 
3 Likes
Same problem…
I’m comparing Flint2 against my old ISP Router/Modem built with legacy Broadcom 63138, 4366 and 43602 chipsets, and my Flint2 is always 6~10dB weaker than it
You’d expect people to know that, but a lot of people will buy the router, set the WiFi channels and/or SSID and then go about their day. And the GL.iNet UI doesn’t even offer a WiFi country code setting.
Maybe someone from Europe who bought the router from Amazon can comment on this, but did your box have French text on it and was the WiFi country code correctly set for your country?
Either way, on Amazon the routers marketed as a “Gaming WiFi Router” with “Rapid OpenVpn & WireGuard” support and “Long Range Coverage”. So it’s not just tech-savvy people who’ll purchase it, which is why I think it’s important for the GL.iNet web UI and the app to guide people into setting the correct country code.
As for this issue, depending on your DNS settings it could be caused by an updated package. But GL.iNet will be releasing an alpha firmware on Monday, so maybe that will fix it?
2 Likes
Cool. Thanks . Looking forward to trying out the alpha build.
The 4.5.7 snapshot 01/28 build does not seem very good. IPv6 is broken. In advanced (LUCI) mode, the WIFI channel scan does not work. Does anyone know the actual MediaTek WIFI driver version in these various builds and can provide that?
I thought GL.iNet would build OpenWrt 23.05 with the MediaTek drivers patched in and then add MediaTek WiFi options to LuCI, since that’s what I’ve seen others do for different routers. We didn’t get that though, so I’d recommend trying an OpenWrt snapshot instead, since that’s what I’ve been using for a month without any issues. Well, apart from the mediocre 2.4GHz range and speed.
If you can’t do without GL.iNet’s web UI then I guess you’re better off using whichever firmware gives you the least amount of issues.
It’s a snapshot. Snapshots are highly unstable. So it’s totally fine.
Repeat after me:
I will not use snapshots and complain about issues.
One more time:
I will not use snapshots and complain about issues.
funny in? I hope they fix the problems and leave the MT6000 in the status of what it should be… top of the line
1 Like
@admon I understand what a snapshot build is. The main issue is not whether or not I used a snapshot build. The issue is lack of progress on correcting the WIFI issues cited in this thread. The “hope” of running the snapshot was to see if later firmware might improve any of the situations. And no it did not fix anything. And yes if you use an OpenWRT build for this router the only one available is snapshot, which is not an issue as generally the snapshot builds have been pretty OK. Much of the WIFI problem cited in this thread is on GL.iNet to get fixes from MediaTek in the form of later drivers. It would be nice if they did that.
3 Likes
Working hard one the new firmware.
6 Likes
When will it be released? Do you have an approximate date? I just started testing the Flint 2 today