In the past, mouse releases from Darmoshark have always been affordable, but rarely convincing when it comes to quality. While the M3 continues to be incredibly affordable, quality is on an entirely different level.

In terms of shape, the M3 isn't particularly original: The M3 is an almost exact recreation of the Finalmouse Air58 shape, and matches it in size, too. However, Darmoshark has managed to hit the same 58 g weight target, despite not having holes and being wireless. In fact, the M3 weighs the same as the Razer Viper V2 Pro, all while being a decent bit larger. Unlike Pulsar with their X2 or LAMZU with their Atlantis, Darmoshark didn't have to resort to tricks such as leaving the bottom shell open to get the weight this low, either. In terms of build quality, the M3 too leaves a good impression: There is only a minor rattle when shaking, no creaking or flexing of the shell when applying lateral force, and even though the side buttons can be actuated by pressing below them, doing so requires excessive force. While there is an indicator LED for the scroll wheel, the M3 doesn't have any RGB lighting, in the interest of both weight and battery life. Speaking of, the M3 comes with a 500 mAh battery, which charges decently fast, and the charging cable is flexible enough to allow playing while charging without feeling overly restricted. The software includes a battery charge indicator, which is reasonably accurate and reliable, and allows me to gauge expected battery life, which I estimate to clear the 60 hours mark in 2.4 GHz operation. The software is a real treat on the M3: It has all the options one would need, I haven't encountered any bugs during testing, and resource usage is minimal. In short: perfect.

My sample is equipped with Kailh GM 8.0 switches for the main buttons, which provide the expected firm and snappy button response. That said, the button design leaves something to be desired: Though lateral movement is minimal, the way the buttons sound and feel changes considerably depending on where they are pressed. To some degree, this is the case on most mice, but the magnitude is higher than average on the M3. A set of Huano switches is used for the side buttons, which is commendable at the price point of the M3, and while the back button suffers from significant post-travel and uneven actuation, the forward button is perfectly fine. For the scroll wheel, an encoder from F-Switch sees use, which scores with nicely separated steps allowing for controlled and tactile scrolling, but noise levels are a bit higher than I'd like them to be. The feet are made of pure PTFE and glide well enough, but on my sample, it seems the base isn't entirely even or the application improper, which is why some scratching is present. Remarkably, a set of replacement feet is included with the mouse, which is nothing short of incredible at this price point.

When it comes to performance, the M3 delivers a surprisingly strong showing. Instead of the nearly ubiquitous CX52850 MCU seen in recent releases from Pulsar, LAMZU, Fantech, Xtrfy, and many others, Darmoshark has opted for a solution from Telink, coupled with PixArt's latest PAW3395 sensor. In my testing, the only major issue I could find relates to CPI deviation, which is ridiculously high. Thankfully, this can be remedied with relative ease, as the deviation is linear, so by lowering nominal CPI accordingly in the software, the desired actual CPI step is easily achieved. Baseline performance is stellar on the M3. In 2.4 GHz operation, wireless motion delay is in the range of 0.5–1 ms, and even enabling MotionSync, which synchronizes SPI timing with USB polling events, only adds at most 0.5 ms of motion delay. The M3 has a wireless extender included, which definitely helps in the latency department as well. With debounce time set to the lowest value in the software, click latency likewise manages to be incredibly competitive, coming out at 1.9 ms in 2.4 GHz operation. Since the lowest debounce time setting elicits occasional slam-clicks, i.e., accidental clicks upon "slamming" or repositioning the mouse, opting for the 1 ms setting may be preferable, which nets a still highly competitive click latency of 2.9 ms. However, there is an issue affecting both motion delay and click latency negatively. For the most part, polling is perfectly stable on the M3. Occasionally, though, singular off-period polls (i.e., 2 ms, 3 ms, 4 ms, etc.) can be observed, which increase both motion delay and click latency. Given that this happens exclusively in 2.4 GHz wireless operation, I'm inclined to believe that dropped packets may be the cause. In regard to motion delay in particular, these off-period polls appear to increase motion delay in such a way that it is maintained throughout the motion. That said, I have no way of knowing this for sure, and while using the mouse in-game, I didn't notice anything, either. Hence, while this does add a layer of inconsistency, I'm inclined to file this under "observable in testing, but negligible in practice." The same goes for some polling instability present in Bluetooth operation, as Bluetooth wouldn't be used for anything latency-critical anyway.

The bottom line for the M3 is pretty straightforward. For an $80 release, the M3 already would be a top contender. For the $49.99 it actually retails (on Mechkeys.com, for instance), the M3 is nothing short of incredible. In terms of performance, the M3 punches way above its weight, and directly competes with many $90 wireless mice, and buttons as well as build are nothing to scoff at, either. Even the Fantech Helios XD3 V2 and Helios Go XD5 for around $60 each can't keep up with the M3 in terms of value. Of course, Darmoshark isn't an established brand, so one may have to lower one's sights when it comes to warranty and support, but given the value proposition brought forward by the M3, I think this is more than acceptable. As such, the M3 earns our Recommended and Budget awards.

When looking at the wireless, ambidextrous competition, there is no shortage whatsoever. The $149.99 Razer Viper V2 Pro has a similar weight, better performance, slightly better build quality, and comes with optical main button switches, along with the option for 4000 Hz wireless polling by virtue of the $29.99 HyperPolling Wireless Dongle, although the software can be a nuisance. The ASUS ROG Harpe Ace Aim Lab Edition weighs even less, has better performance, similarly good buttons and build quality, but Armoury Crate can be headache-inducing, for $149.99. For $129.00, the G-Wolves Hati-S Plus ACE weighs less while lacking much of a bottom plate and having a smaller battery, has largely similar performance, a similarly good build and buttons, and a stiffer charging cable. The $119.99 VAXEE XE Wireless has largely comparable performance, a higher weight, better buttons and build, and extensive on-board configurability. For $99.99, the Xtrfy M8 Wireless comes with similar performance, similar weight and also no holes, great buttons and build, and no need for software. The $94.95 Pulsar X2 and X2 Mini have similar performance, a slightly lower weight due to lacking a proper bottom plate, and similar button quality, but some issues with quality control. For $89.99, the LAMZU Atlantis and Atlantis Mini likewise have similar performance, a lower weight due to lacking a proper bottom plate, and similar button quality. At 68 g, the Fnatic BOLT weighs significantly more, struggles with its buggy software and dull side buttons, but scores with good performance, for $89.99. For $79.99, the Endgame Gear XM2we has similar performance, a similar weight while too lacking holes, optical switches for the main buttons, and perfectly functional software. Lastly, the Cooler Master MM712 is the budget option at $69.99, with slightly worse performance on the latest firmware, optical main button switches, a similar weight, and a poor side button.