The Model O 2 Mini Wireless is a very straightforward product. As the name already implies, it is a smaller Model O 2 Wireless, and thus the direct successor of the Model O- Wireless. However, due to the internals and firmware carrying over as well, the Model O 2 Mini Wireless has inherited all of the weaknesses its larger sibling already suffered from.

Being 6% smaller than the Model O 2 Wireless, the Model O 2 Mini Wireless weighing less comes as no surprise. That said, at 57 g, it is 14 g lighter, which is more than the difference in size would suggest, especially given the fact that both the overall design and battery capacity are the same. In fact, the Model O 2 Mini Wireless manages to be 1 g lighter than the trusty old Model O-, which was wired. Despite this weight reduction, the Model O 2 Mini Wireless is easily as well-built as the Model O 2 Wireless. On my sample, there is no rattle when shaking the mouse, no creaking or flexing of the shell when applying lateral pressure, and the side buttons cannot be actuated by pressing below them regardless of force used. No changes have been made to the feet, either, which still are made of pure PTFE and glide well, and the cable likewise is unchanged, offering up great flexibility to where playing while charging feels indistinguishable from using the wired Model O 2 Mini.

Much like the Model O/I 2 Wireless, the Model O 2 Mini Wireless utilizes Kailh GM 8.0 switches for the main buttons, which provide firm and snappy actuation, moderate pre and post-travel notwithstanding. Some lateral button movement is present, but still low enough to only become apparent when actively provoking it. The side buttons are outfitted with 2-pin low-profile switches from Kailh, which shine with very low post-travel, although pre-travel is low to moderate. Due to their size, feedback isn't particularly pleasing, though the actuation point is even across their entirety at least. For the scroll wheel, an encoder from F-Switch (yellow, green core) sees use, which is familiar from the Model O/I 2 Wireless, and delivers good tactility at low noise levels.

In short, as far as the physical part is concerned, the Model O 2 Mini Wireless does well. When it comes to performance, on the other hand, things are looking rather grim. Much like on the Model O/I 2 Wireless, Glorious has opted for an MCU from PixArt, along with PixArt's PAW3395 sensor. A similar MCU has been used on VAXEE's wireless releases, which shined with great performance, but on the Model O 2 Mini Wireless, we're seeing the exact opposite. First of all, while not overly large, CPI deviation is highly inconsistent, which is relevant in that instead of knowing that nominal and actual steps will deviate by the same percentage no matter what, and therefore being able to account for it quite easily, one has to measure each step to determine the actual deviation, adding another layer of inconvenience. A recurring issue from the Model O/I 2 Wireless is polling stability, or rather the lack thereof. Put simply, there isn't a single polling rate not suffering from various amounts of instability, and disabling RGB lighting does nothing in this regard, either. In addition, the tracking irregularities observed on the Model O/I 2 Wireless also make a return, especially the one whereby, when moving from a standstill, counts registered by the sensor are suspended for a short time, and upon being resumed are reported in a single update. As a result, responsiveness suffers, and in addition, there is also the issue that at any point during the motion, motion delay randomly increases by a multiple of the set interval. Taken together, we're seeing an above average degree of inconsistency, to where I actually found myself struggling to determine the exact motion delay in testing. Under ideal conditions, a motion delay differential of 1 ms without MotionSync and 2 ms with MotionSync relative to the Logitech G403 (control subject) can be expected, but in actual usage, rather expect this to be all over the place.

The story continues with click latency. Again, much like on the Model O/I 2 Wireless, results will vary greatly depending on what the interval between each sample is. At 200 ms or below, the numbers are quite formidable: at a debounce time of 0 ms, 2.4 ms are averaged in wireless (2.4 GHz) operation. However, as soon as the interval is increased to 300 ms or more, the tides turn, and the average now is 7.4 ms, with a near-absurd standard deviation of 3.44 ms. Essentially, what is happening here is that in order to save battery life, the RF signal is suspended whenever there is no sensor and button activity for more than 200 milliseconds. Upon being resumed, various amounts of latency are added, resulting in a worse average and huge standard deviation. Of course, during actual use, there will be sensor activity during clicks most of the time, which is why the former value applies, but when playing games such as CS2 or VALORANT, it is not uncommon at all to not move the mouse for 300 ms, for instance when holding an angle. In these cases, the latency of the next click latency can end up anywhere between 2 and 12 ms, adding a layer of inconsistency that is essentially impossible to account for. Thus, I consider the Model O 2 Mini Wireless not suited for competitive gaming.

While the software for the Model O 2 Mini Wireless is still Glorious Core, it has received a large overhaul with version 2.0, mostly on the visual and functional level. If anything, this overhaul is at least a step back, as not only did resource usage increase again, the UI has also turned to the worse. The biggest offender in this regard is undoubtedly CPI adjustment. Whenever a step is set, Core no longer permits changing it. Instead, one has to delete the step and add a new, modified one, and since new steps are always added sequentially, and aren't ordered automatically, or can be ordered manually, one also has to delete all other steps potentially following the one in need of adjustment. Furthermore, the colors the steps are coded with cannot be modified either. In short, CPI adjustment is realized in the most backward way possible, and this is doubly annoying given that all of this was possible before the 2.0 overhaul. In general, navigating Core hasn't gotten any easier with the overhaul, and coupled with increased installation size and resource usage, the fact that cloud profile storage now asks for a Glorious ID can be filed under minor annoyance.

Much like before, Core also includes a battery life indicator, which is percentage-based and features seemingly single-digit accuracy. In practice, however, the indicator did not budge at all for me, even when using the most draining settings (1000 Hz, illumination enabled with 100% brightness) for several hours. After plugging the mouse in once and then going back to wireless, a non-100% value was shown, but every time this was done the value went up or down, so its reliability appears limited at best. In short, the only numbers we can go by are those cited by Glorious, which are 110 hours using 2.4 GHz and 210 hours using Bluetooth, both without illumination. With illumination, these numbers will be much lower, but by how much is anyone's guess. Charging at least is decently speedy, which, coupled with the flexible charging cable, should make for a largely seamless experience.

I've always had a soft spot for the Model O- shape, but unfortunately, the Model O 2 Mini Wireless falls short in the performance department, and doesn't offer enough otherwise to even come close to justifying its $99.99 price tag. Small, ambidextrous mice aren't exactly a rare commodity, and the alternatives offered by Pulsar (X2V2, $99.95), Endgame Gear (OP1we, $89.99), or Fantech (Helios II Pro, $74.80) all run circles around the Model O 2 Mini Wireless. In fact, while sporting a different shape, Glorious's own Series 2 Pro Wireless is very much worth the $30 premium over the Model O 2 Mini Wireless, too.