When it released back in 2019, the Glorious Model O made a big splash on the market: Priced at $49.99, it popularized the concept of a lightweight gaming mouse by virtue of being much more affordable than the offerings from Finalmouse. More than four years later, Glorious gives a thorough update to the mouse that made them a name in the industry. Unfortunately, the Model O 2 suffers from several issues that render it a less compelling offer than the original version.

The original Model O was among the first mice that pioneered the generous use of holes in the mouse shell to lower weight. The Model O 2 continues to employ this technique, although the holes are circular instead of hexagonal now, and the weight is lower at 57 g, down from 68 g. That said, a lot has happened in the mouse industry in the past four years, and the market has shifted towards solid-shell designs again, either with a closed or open bottom. Accordingly, 57 g on a wired mouse with RGB lighting and plenty of holes on the top and bottom isn't anything special anymore, and if anything, I'd argue that most people would rather have a 61 g mouse with a solid shell than a 57 g mouse with holes all over. Moreover, compared to a release such as the solid-shell Darmoshark M3, which is an even larger mouse, yet weighs 58 g while being wireless, although without RGB lighting, the Model O 2 compares even less favorably. Build quality at least is fine on my copy: No rattle when shaking, no creaking or flexing of the shell when applying lateral pressure, and no side button actuation when squeezing below them.

The buttons are of generally good quality on the Model O 2. Much like other recent Glorious mouse releases, the Model O 2 is outfitted with "Glorious switches," which in fact are modified Kailh GM 8.0. Button response is accordingly firm and snappy, though quite a bit stiffer than it was the case with the Omron D2FC-F-7N (20M) (OF) switches used on the first-generation Model O. The button design has been revised as well, as the main buttons no longer extend across the entire width and instead are framed by ledges. Lateral button movement is minimal in any case, which wasn't always true on the Model O and thus is an improvement. For the side buttons, a set of Huano switches sees use, with mixed results. While pre and post-travel are under control on the forward button, the back button has noticeable pre-travel on my copy, resulting in a rather mushy response. For the scroll wheel, Glorious once again has opted for an encoder from F-Switch, which I've always been a fan of, and results are indeed excellent: Noise levels are low, and tactility is great, providing clearly separated steps allowing for controlled scrolling. I have no complaints about the feet, either, which are made of pure PTFE and glide very well. In addition, their size and shape is unchanged compared to the original Model O, allowing users to re-use third-party skates made for the latter. Lastly, the cable is pretty much unchanged to other recent Glorious wired mouse releases and nicely flexible, though the cable of the HyperX Pulsefire Haste 1/2, for instance, still has it beat by quite a margin.

Performance is where the Model O 2 is starting to falter. PixArt's PAW3395 sensor is used, which is paired with a rather obscure and rebranded MCU from Westberry. Still, there is no reason not to expect flawless performance in a wired mouse. This isn't the case, however. For one, CPI deviation is significant on the Model O 2, and more importantly, it isn't consistent, either. Another issue relates to onset motion delay. Being primarily conceived for wireless operation, the 3395 has several power-saving techniques, which can result in a slower sensor framerate ramp-up when moving the mouse from a standstill. Obviously, saving power is pointless on a wired mouse, yet the Model O 2 does exactly that, leading to increased motion delay at the onset of motion. Motion delay is also negatively affected by MotionSync, which has been touted as a stand-out novelty by Glorious. On paper, it synchronizes SPI reads with USB polling events to facilitate more consistent timing between the two, but on the Model O 2, this also increases motion delay by close to 1 ms at 1000 Hz, which isn't really what I would call a selling point. Luckily, it can be turned off, but when doing so, there really isn't any meaningful difference in tracking performance between the trusty old 3360 used in the original Model O and the new 3395. Another curious point is lift-off distance, as contrary to virtually every other mouse featuring this sensor, lift-off distance at the lowest setting is above 1 DVD. Hence, those expecting LOD similarly low to competitors likewise using the 3395 won't find it on the Model O 2. Lastly, the only fully stable polling rate on the Model O 2 is 1000 Hz, as all polling rates below it suffer from periodic off-period polls.

Click latency isn't too exciting on the Model O 2, either. If set to the default debounce time of 10 ms, click latency is unacceptably high, which is why lowering it to 0 ms in the software is pretty much a necessity. After doing so, click latency sits at 2.6 ms, which is very good, but since defer-type debouncing is used, the Model O 2 already slam-clicks at this setting, which describes inadvertent button actuation when "slamming" or gently repositioning the mouse on the pad after lift-off, and eventually, double-clicking may follow suit. To compensate, one has to increase debounce time again, in turn increasing click latency to less competitive values. Speaking of the software, the Model O 2 of course is compatible with Glorious Core, which is largely bug-free, but continues to have quite the appetite for CPU time in particular. Thankfully, a recent update has at least quelled its hunger for RAM, which is now only high instead of excessively high.

Overall, the Model O 2 is not a bad mouse, just an unimpressive one. The real issue is the pricing. Compared to a competitor such as the HyperX Pulsefire Haste 2, which comes with 8000 Hz polling, is lighter, and has no holes, paying $5 more for the Model O 2 to get virtually nothing better in return for is difficult to argue. In fact, I have a hard time making a case against simply getting the original Model O instead, which doesn't perform any worse and is infinitely more affordable. With the market largely having shifted to wireless, selling a wired mouse in 2023 is only feasible if it excels in performance (like the Razer DeathAdder V3) or is sold at a price that makes it interesting for budget-minded buyers, and neither applies to the Model O 2.