Apex Neural Rendering vs NVIDIA DLSS Frame Generation: Can ANR Become a Serious Rival?
AI-powered frame generation is quickly becoming one of the biggest performance-boosting technologies in PC gaming. Instead of relying only on frames rendered directly by the game engine, frame generation creates additional AI-generated frames and places them between the traditionally rendered ones. The result can be much smoother motion, higher displayed frame rates, and a more fluid gaming experience, especially in demanding titles with heavy ray tracing or advanced visual effects.
NVIDIA DLSS Frame Generation is currently one of the best-known solutions in this space, with Multi Frame Generation taking the concept even further by inserting multiple generated frames between rendered frames. Intel XeSS Frame Generation and AMD FSR Frame Generation are also part of this growing trend. But as useful as frame generation can be, it is not perfect.
Generated frames can introduce visual artifacts, ghosting, flickering, uneven frame pacing, added latency, and extra processing overhead. These problems can become more noticeable when moving from standard 2X frame generation to more aggressive 3X or 4X Multi Frame Generation modes.
That is where Apex Neural Rendering, or ANR, enters the picture.
Developed by Apex Compute, Apex Neural Rendering is an emerging AI rendering suite that aims to compete with established technologies like NVIDIA DLSS. Its most interesting feature right now is ANR Frame Generation, which is already available in beta and includes support for Multi Frame Generation.
The big question is simple: can Apex Neural Rendering challenge NVIDIA DLSS Frame Generation in real games?
What makes Apex Neural Rendering interesting?
Apex Neural Rendering is designed as a full AI-powered rendering package made up of three main technologies: Frame Generation, Super Resolution, and Ray Reconstruction.
At the moment, only ANR Frame Generation is publicly available in beta form. Super Resolution and Ray Reconstruction are expected to arrive later, which could eventually make ANR a more complete alternative to existing AI upscaling and rendering platforms.
One of ANR’s biggest selling points is hardware compatibility. While NVIDIA’s most advanced frame generation features are officially limited to newer RTX graphics cards, ANR is being positioned as a broader solution. It is expected to work on GeForce GTX 16 Series and RTX 20 Series GPUs and newer, Radeon RX 5000 Series GPUs and newer, and Intel Arc graphics cards.
That broader compatibility could make ANR especially appealing to PC gamers who own older but still capable GPUs. If the technology delivers solid image quality and stable performance, it could give many users access to advanced frame generation without needing a latest-generation graphics card.
Apex Compute also claims that ANR Frame Generation uses significantly less GPU memory than NVIDIA DLSS Frame Generation. According to the company’s own benchmark figures, ANR FG consumes around 214 MB of VRAM at 4K, compared with around 562 MB for DLSS FG. That would represent a reduction of roughly 62 percent.
The developer also claims faster frame-generation processing in selected workloads, along with improved scores in several image-quality measurements. However, internal benchmarks do not always translate directly into real-world gaming results. Faster frame generation processing does not automatically mean higher FPS in every game, and image quality can vary depending on the title, scene, camera motion, and frame generation multiplier.
That is why real gameplay testing is essential.
How ANR Frame Generation is installed
ANR Frame Generation is distributed as a beta package from Apex Compute’s official development source. For testing, the ANR FG V.B15 release is used. This version is notable because it fixes a VRAM leak that could happen when games repeatedly reinitialized Frame Generation.
The full package includes OptiScaler, so a separate OptiScaler installation is not required. After extracting the archive, the game should be closed before running the included installation script. The installer is pointed toward the folder containing the game’s executable file.
ANR works by integrating through OptiScaler and intercepting the game’s existing Frame Generation implementation. Once installed, the player enables DLSS Frame Generation from the game’s graphics menu, and ANR automatically takes over in the background. No separate in-game toggle or hotkey is required.
Users can confirm that ANR has initialized correctly by checking the generated log files in the game’s OptiScaler custom folder for Apex Neural Rendering FG. If removal is needed, the included uninstall script can be run against the same game directory.
Because ANR is still in beta, users should expect the possibility of bugs, compatibility issues, and inconsistent behavior depending on the game.
Testing Multi Frame Generation on older RTX hardware
The test system uses a GeForce RTX 4060 Laptop GPU. Officially, this GPU supports standard 2X DLSS Frame Generation, but not the higher Multi Frame Generation multipliers associated with newer hardware.
To compare 3X and 4X configurations, a community-made Universal MFG Unlocker is used. This enables higher DLSS Multi Frame Generation multipliers on compatible older GeForce GPUs, including RTX 40 Series hardware.
Previous independent testing has shown that DLSS Frame Generation can behave very similarly on modified older GPUs compared with newer models, including comparable interpolation artifacts and frame pacing. However, performance and latency can still vary depending on the GPU, driver, game engine, and implementation.
For this comparison, the goal is to test matching frame generation multipliers between NVIDIA DLSS and ANR. That means comparing 2X against 2X, 3X against 3X, and 4X against 4X where possible.
Test system and methodology
All testing is performed on an ASUS TUF Gaming F15 laptop with an Intel Core i7-12700H processor, a GeForce RTX 4060 Laptop GPU with 8 GB of GDDR6 memory and up to 140 W power, 16 GB of dual-channel DDR4-3200 memory, and Windows 11. The test resolution is 1920 x 1080.
Performance analysis is handled through CapFrameX 1.8.1, while gameplay footage is captured using NVIDIA’s recording tools.
For frame generation testing, standard game FPS readings are not enough. Frame generation can increase the number of displayed frames, but the game engine still processes simulation, physics, and player input at the base rendered frame rate. Because of that, the displayed frame rate and the actual input responsiveness are not always the same thing.
To better evaluate what players actually see on screen, the comparison uses CapFrameX’s Display Times metric, specifically MsBetweenDisplayChange. This helps measure displayed frame delivery rather than only traditional engine-side frame times.
The analysis looks at average displayed FPS, 1 percent lows, 0.1 percent lows, stuttering behavior, and experimental PC latency figures. Each benchmark capture is around 20 seconds long, so the results should be treated as indicative rather than final statistical proof. Short captures can show trends, but they may not fully represent rare stutters or long-session stability.
The games used for testing are Valor Mortis, Cyberpunk 2077, and The Witcher 3: Wild Hunt Remastered.
Valor Mortis is tested with DLSS FG 2X against ANR FG 2X. Cyberpunk 2077 is tested with DLSS MFG 3X against ANR MFG 3X. The Witcher 3: Wild Hunt Remastered is tested with DLSS MFG 4X against ANR MFG 4X.
Additional visual comparisons at 2X, 3X, and 4X help evaluate image quality differences, though video compression can make fine artifacts harder to see. Compression can blur small details, hide subtle ghosting, or create artifacts that are not present during live gameplay, so direct gameplay observation remains important.
Valor Mortis: DLSS FG 2X vs ANR FG 2X
Valor Mortis is a first-person action RPG from One More Level, the studio known for the Ghostrunner series. It is a useful test case because fast camera movement, close-range combat, and quick directional changes can expose weaknesses in frame interpolation.
At 2X Frame Generation, ANR makes a strong first impression. Motion interpolation appears surprisingly close to NVIDIA DLSS Frame Generation, with both solutions producing fluid movement during normal gameplay. In the tested scenes, ANR’s generated frames look convincing enough that many players may not immediately notice a major difference without close side-by-side inspection.
This is important because 2X frame generation is the most common and practical use case for many gamers. If ANR can remain visually competitive at this level, it could become a valuable option for users on hardware that does not officially support newer frame generation features.
As with all frame generation technologies, visual quality depends heavily on the scene. Fast motion, transparent effects, particle-heavy moments, UI elements, and rapid camera turns are usually the hardest areas for interpolation. These are the kinds of moments where flickering, ghosting, or unstable edges can become more visible.
Still, ANR’s early beta performance in Valor Mortis suggests that the technology has real potential.
Why Cyberpunk 2077 and The Witcher 3 matter
Cyberpunk 2077 is one of the most demanding modern PC games and a major showcase for advanced rendering features. It is a strong test for 3X Multi Frame Generation because the game can place heavy pressure on the GPU, especially with ray tracing enabled. In this kind of environment, frame generation can dramatically improve perceived smoothness, but it can also magnify latency and artifact concerns.
The Witcher 3: Wild Hunt Remastered is another valuable test case, particularly at 4X Multi Frame Generation. The remastered version includes modern visual upgrades, improved lighting, and demanding rendering features. Testing 4X generation here helps show how well ANR and DLSS handle more aggressive frame insertion.
Higher multipliers are where frame generation becomes more challenging. At 3X or 4X, the system is generating more artificial frames between traditionally rendered frames. This can produce very high displayed FPS, but it also increases the importance of frame pacing, artifact control, and latency management.
If ANR can remain stable and visually clean at these higher multipliers, it would strengthen its case as a serious alternative to DLSS Frame Generation.
The bigger picture for PC gaming
Apex Neural Rendering is still early, and ANR Frame Generation remains a beta product. It should not yet be treated as a guaranteed replacement for NVIDIA DLSS Frame Generation. DLSS has years of development behind it, wide game support, and strong integration across many major PC releases.
However, ANR is interesting because it targets some of the biggest pain points in current frame generation technology: hardware restrictions, VRAM usage, and processing cost.
If ANR can deliver good image quality, broad GPU support, low memory usage, and stable frame pacing, it could become a major tool for PC gamers who want smoother performance without upgrading their graphics card. This is especially relevant as modern games continue to demand more GPU power and more VRAM.
For now, the most sensible view is cautious optimism. ANR is not perfect, and beta software always comes with limitations. But the early signs are promising enough to make it worth watching closely.
NVIDIA DLSS Frame Generation remains the established solution, but Apex Neural Rendering may become one of the most compelling new competitors in AI frame generation. For gamers using older GeForce, Radeon, or Intel Arc hardware, ANR could eventually offer a practical path to smoother gameplay and higher displayed frame rates across demanding PC titles.Apex Compute’s ANR frame generation shows promise, but NVIDIA DLSS still delivers the smoother experience
Apex Compute’s ANR frame generation is shaping up to be one of the more interesting challengers to NVIDIA DLSS Frame Generation, especially as more games begin pushing beyond simple 2X frame insertion into Multi Frame Generation. In early comparisons, ANR can produce higher average frame rates and even lower reported latency in some scenarios. However, raw FPS is only part of the story. Frame pacing, visual stability, stuttering, UI behavior, and motion fluidity all play a major role in how smooth a game actually feels.
In Valor Mortis, the comparison between NVIDIA DLSS FG 2X and ANR FG 2X was relatively encouraging for Apex Compute’s technology. Visual artifacts were generally well controlled, and the differences between the two frame generation methods were not especially dramatic during gameplay. That said, Valor Mortis may represent a best-case scenario for ANR. At 2X frame generation, only one generated frame is inserted between traditionally rendered frames, which can help hide temporal instability and interpolation issues that become much easier to notice at higher Multi Frame Generation levels.
The performance results in Valor Mortis reveal a clear trade-off. ANR FG reached an average of 170.6 FPS, compared to 153.7 FPS for NVIDIA DLSS FG. That gives ANR an impressive 11% lead in average displayed frame rate. On paper, that looks like a strong win.
However, the lower-percentile performance tells a different story. NVIDIA DLSS delivered 54.7 FPS in the 1% lows, while ANR dropped to 41.3 FPS. The gap became even larger in the 0.1% lows, where DLSS managed 33.2 FPS compared to just 18.5 FPS for ANR. That is a major difference, because 1% and 0.1% lows often reveal the stutters and dips that players actually feel during gameplay.
Frame pacing also favored NVIDIA. ANR showed more severe frame time spikes and spent about 3.5% of the test classified as stuttering, compared with 1.8% for DLSS. Interestingly, reported PC latency was very close, with ANR slightly ahead at 35.5 ms versus 37.5 ms for NVIDIA.
The takeaway from Valor Mortis is that ANR can deliver higher average FPS, but DLSS still offers better consistency. Even so, the visual comparison felt closer than the numbers suggest, with both technologies appearing fairly comparable in perceived motion fluidity during this particular test.
Cyberpunk 2077 proved to be a much tougher challenge. As one of the most visually demanding PC games, it creates the kind of environment where frame generation weaknesses are easier to expose. Dense city streets, fast camera movement, traffic, animated NPCs, advanced lighting, and ray tracing all give interpolation algorithms plenty of opportunities to fail.
For this test, both technologies were compared using 3X Multi Frame Generation, meaning two generated frames were inserted between traditionally rendered frames.
This is where ANR began to struggle more noticeably. Compared with NVIDIA DLSS MFG, ANR showed more flickering, weaker temporal stability, and a higher number of interpolation artifacts, especially during camera movement. Motion also felt more jittery and less consistent, reducing the benefit of the higher displayed frame rate.
Another significant issue was the HUD and UI behavior. In the tested configuration, ANR did not interpolate HUD and interface elements. As a result, the game world appeared to update at a higher generated frame rate, while the UI continued refreshing at the lower base rendering frame rate. At 3X Multi Frame Generation, that mismatch became very noticeable and made the overall presentation feel less polished.
From a performance perspective, Cyberpunk 2077 looked much closer. ANR averaged 224.1 FPS, while NVIDIA DLSS averaged 218.8 FPS. That gives ANR a small 2.4% lead in average frame rate. The 1% lows were nearly identical, with ANR at 108.3 FPS and DLSS at 109.2 FPS.
NVIDIA still held an advantage in the 0.1% lows, reaching 73.3 FPS compared with 64.7 FPS for ANR. Stuttering was minimal for both solutions, with DLSS at 0.2% and ANR at 0.3%. ANR also reported slightly lower PC latency at 41.8 ms, compared with 45.3 ms for DLSS.
Looking only at the benchmark numbers, Cyberpunk 2077 appears almost like a draw. ANR offers a slightly higher average FPS and lower reported latency, while DLSS provides somewhat stronger minimum frame consistency. But actual gameplay tells a more important story. NVIDIA DLSS MFG looks and feels more stable, with smoother motion and fewer distracting artifacts. In this case, visual quality matters more than the small FPS advantage.
The Witcher 3: Wild Hunt Remastered offered the most interesting comparison. With its expansive forests, dense foliage, horseback traversal, detailed character animation, and upgraded lighting, it provides a demanding workload for any frame generation solution. The remastered version introduced path tracing, NVIDIA DLSS 4.5, improved vegetation, enhanced lighting, and other visual upgrades that make it a strong test case for Multi Frame Generation.
This time, both technologies were tested at 4X Multi Frame Generation, inserting three generated frames between each pair of traditionally rendered frames.
The results were mixed. ANR did a surprisingly good job eliminating some of the distracting screen-edge tearing and wavy distortion that could appear with NVIDIA DLSS MFG. This was especially noticeable during camera movement around foliage and fine environmental details. In those moments, ANR produced a cleaner image in areas where DLSS could show visible distortion.
However, ANR still struggled with perceived smoothness. Motion interpolation felt less fluid and more jittery compared with NVIDIA’s implementation. The generated frames did not always feel evenly paced, which reduced the sense of smoothness that 4X Multi Frame Generation is supposed to provide.
The HUD and UI issue also became even more distracting at 4X. Since the interface elements were not interpolated in the tested ANR setup, they updated at the much lower base rendering frame rate while the game world ran at a far higher displayed frame rate. The larger the Multi Frame Generation multiplier, the more obvious this mismatch becomes.
Performance results in The Witcher 3 were fascinating. ANR averaged 277.5 FPS, while NVIDIA DLSS averaged 259.9 FPS. That gives ANR a 6.8% lead in average displayed frame rate. NVIDIA maintained stronger 1% lows at 162.5 FPS compared with 151.7 FPS for ANR. However, ANR reversed the situation in the 0.1% lows, reaching 75.0 FPS compared with 56.7 FPS for DLSS.
Stuttering was low for both, with ANR at 0.3% and DLSS at 0.4%. In this specific metric, ANR had a slight advantage. But latency was a major weakness. ANR reported 57.8 ms of PC latency, while NVIDIA DLSS came in much lower at 38.6 ms. That is a large difference and could be noticeable in gameplay, particularly in a game where camera control, movement, and combat responsiveness matter.
Across all three games, Apex Compute’s ANR frame generation shows real potential. It can deliver higher average FPS than NVIDIA DLSS Frame Generation in several scenarios, and in some cases it even reduces certain visual artifacts. The Witcher 3 comparison is a good example, where ANR cleaned up some screen-edge distortion that affected DLSS.
But the broader picture still favors NVIDIA DLSS, especially when looking beyond average frame rate. DLSS generally provides stronger frame pacing, better temporal stability, smoother motion, and more consistent lower-percentile performance. It also handles the overall gameplay presentation more convincingly, particularly in Cyberpunk 2077, where ANR’s flickering, jitter, and non-interpolated UI elements became difficult to ignore.
ANR is a promising start, but it is not yet a clear DLSS replacement. Higher FPS numbers are appealing, but smoothness depends on more than average performance. Until ANR improves its temporal stability, HUD interpolation, motion consistency, and latency behavior at higher Multi Frame Generation levels, NVIDIA DLSS remains the more polished and reliable option for players who care about both image quality and fluid gameplay.Apex Neural Rendering vs NVIDIA DLSS: Higher FPS Does Not Always Mean Smoother Gameplay
Apex Neural Rendering is shaping up to be an ambitious frame generation technology, and early testing suggests it has real potential. Across several games, ANR was able to produce higher average displayed frame rates than NVIDIA DLSS Frame Generation and Multi Frame Generation. On paper, that sounds like a major win. In practice, however, the experience is more complicated.
The biggest takeaway is that raw FPS numbers do not tell the full story. A game can show a higher frame rate counter while still feeling less smooth, less responsive, or less visually stable. That is exactly where ANR currently struggles.
In Valor Mortis, Apex Neural Rendering delivers its best showing. At 2X frame generation, its motion interpolation quality comes surprisingly close to NVIDIA DLSS Frame Generation. The image remains relatively stable, and the overall presentation is competitive enough to show that the technology has a strong foundation. Even so, the weaker low-percentile performance suggests that frame pacing is not yet as consistent as it needs to be.
The situation becomes more challenging when higher frame generation multipliers are used. In Cyberpunk 2077, ANR Multi Frame Generation produces higher average FPS, but the visual trade-offs are much easier to notice. Excessive flickering, weaker temporal stability, and jitterier motion make gameplay feel less refined. Instead of delivering a fluid high-frame-rate experience, the added frames can sometimes call attention to themselves.
The Witcher 3: Wild Hunt — Remastered presents a more mixed result. ANR does a better job than NVIDIA’s solution in one important area: it reduces the distracting screen-edge distortion that can appear with DLSS Frame Generation. That is a meaningful advantage, especially in a third-person open-world game where camera movement and edge detail are constantly visible.
However, that improvement is offset by several issues. The perceived motion smoothness is weaker, HUD elements are not interpolated properly, and responsiveness takes a noticeable hit. Reported PC latency measured around 57.8 ms with ANR, compared to 38.6 ms for NVIDIA DLSS Multi Frame Generation. That is an increase of 19.2 ms, or almost 50%.
While experimental latency measurements should always be treated with caution until they are verified through additional testing, the gap is large enough to matter. In fast-paced games, higher latency can make controls feel heavier and less immediate, even if the frame rate counter looks impressive.
This is the key problem for Apex Neural Rendering right now. It can boost average displayed FPS, and in some cases it can produce better 0.1% lows, but those gains do not automatically translate into a better gaming experience. Smooth motion, consistent frame delivery, clean image quality, and low input latency are just as important as the final FPS number.
At 3X and 4X frame generation, ANR’s weaknesses become more visible. Inconsistent interpolation quality, unstable motion, flickering, and less polished frame pacing prevent it from matching the overall feel of NVIDIA DLSS. The technology is promising, but it is not yet as mature.
That said, Apex Neural Rendering deserves attention. Its broad hardware compatibility could make it attractive to many PC gamers, especially those who do not own the latest graphics cards. Its ability to reduce certain screen-edge artifacts also shows that it can solve problems that still affect competing frame generation solutions.
For now, NVIDIA DLSS Frame Generation and Multi Frame Generation remain the more polished options overall. They generally deliver better motion stability, smoother perceived gameplay, and lower latency. Apex Neural Rendering can win in average FPS, but NVIDIA’s solution currently provides the more balanced and refined experience.
The good news is that ANR is still in beta. With further optimization, improved temporal stability, better HUD handling, and reduced latency, it could become a much stronger competitor. Upcoming additions such as Super Resolution and Ray Reconstruction may also help Apex Compute build a more complete image-enhancement ecosystem.
Apex Neural Rendering is not ready to dethrone NVIDIA DLSS yet, but it is a technology worth watching. If future updates can turn its high frame rate numbers into genuinely smoother and more responsive gameplay, it could become a serious option for PC gamers looking for broader frame generation support.
