explainx.ai350K+ learn from us
TrendingAI News TodayPathwaysSkills
Pricing
explainx.ai

Upskill in AI — 16 free pathways, live workshops & bootcamps, and 50+ courses from practitioners. Plus the skills, tools, and MCP servers to practice on.

follow us

follow on google

Add explainx.ai as a preferred source

corporate training

support@explainx.ai

get started

Find your pathTake Free Evaluation

community

Join the community

learn

mind: share how you thinkpathways — start freeworkshopsbootcampscoursescompare Explainxcertificationsmock testsexplainx universitycorporate traininglearn skills & mcp

discover

skillsmcp serversexplainx mcpestacktoolsmdx readeragentsllmsdesignsdictionarypeopleagi trackerfelony benchranks

company

aboutvisionmissionteaminstructorsteach on explainxpartnershipscommunityhackathonscareers

content

daily AI newsstate of AI — live resultsblogreleasespromptsgeneratorsresource libraryfor LLMsexplainx.ai kids

solutions

all solutionsdeveloper upskillingmarketing upskillingproduct manager upskillingleadership upskilling

newsletter · weekly

Get AI news, tools, and insights in your inbox.

supportcontactprivacytermsdata rightshow we create contentsubmission guidelines

© 2026 AISOLO Technologies Pvt Ltd

explainx.ai

On this page

  • TL;DR: claimed vs verified
  • What is Veyra?
  • Why the GTA6 and DLSS 5 tagline is a problem
  • How do the sources work?
  • What are the limits the author lists?
  • What are the licensing and trust questions?
  • How should you try it safely?
  • How does it compare with cloud video upscalers?
  • What are people asking?
  • What this means for you
  • Related reading
← Back to blog

explainx / blog

Veyra: The GPL Windows Video Enhancer Claiming PS5 GTA 6 at 120 FPS With DLSS 5, Claimed vs Verified

Open Source, GitHub Trending, NVIDIA, Video AI, Claimed vs Verified

Part of AI Tools and Apps

Veyra (Veyra-NRVideo) is a GPL-3.0 Windows tool for AI upscaling, denoising and frame generation. What its README proves, what it does not, and the risks.

Nov 13, 2026·8 min read·Yash Thakker
add explainx.ai
go deep
Veyra: The GPL Windows Video Enhancer Claiming PS5 GTA 6 at 120 FPS With DLSS 5, Claimed vs Verified

A GitHub repo whose tagline says "Bring PS5 GTA6 to 120FPS with DLSS 5" will get attention, and Veyra-NRVideo did: it was trending in the week of November 9, 2026, and the page now shows about 1,400 stars and 65 forks. The tool is real and actively maintained. The tagline is the part we could not verify.

This post separates what the README documents from what the tagline implies, explains how the tool works, and lists the risks before you download a GPU-heavy Windows binary from a project hosted under an individual's GitHub account.

TL;DR: claimed vs verified

table · 3 cols
ClaimWhat the README showsVerdict
"Bring PS5 GTA6 to 120FPS with DLSS 5" (repo tagline)No GTA6 test or 120 FPS measurement in the README bodyNot verified
Real-time AI super resolution, denoising, HDR, frame generationDetailed feature lists, version history to 2.1.5, cost examples on an RTX 5070Documented by the author
PS5, PC and Xbox streamingPS5 Remote Play, Sunshine PC streaming, Xbox marked unofficial and experimentalDocumented, with caveats
"DLSS 5"README: community runtimes are not vendor certification or a complete official DLSS 5 integrationAuthor hedges the tagline
79K+ portable-package downloadsAuthor-reported as of 2026-11-08, counts are download events, not unique usersAuthor-reported
Open sourceOriginal code GPL-3.0; source excludes proprietary runtimes and modelsPartly open
Daily briefFrom Yash · 350K+ students taught

Catch up on AI

One AI story explained in plain English, plus one thing to try. Every day, free. Unsubscribe anytime.

What is Veyra?

Veyra calls itself "Real-time GPU Video Enhancement for Windows." The README's one-line pitch: turn local video, capture cards and game streaming into one real-time GPU enhancement pipeline. It is a C++ app with a Qt Quick interface and a native Direct3D 12 video window, for Windows 11 on NVIDIA and AMD GPUs.

The feature set, per the README:

  • Super resolution (SR): NVIDIA RTX Video SR families, DLSS, XeSS and FSR options, with true dual SR and SR-then-downsample added in version 2.1.2.
  • Noise reduction (NR): up to four independent NR layers, with model choices such as SF310.9.16 for RTX 20 to 50, NVIDIA original for RTX 50, and experimental AMD models for RX 7000 and 9000.
  • Frame generation (FG): a choice of backends and multipliers, with checks for cadence and display sync.
  • Colour and HDR: curves, mixer, wheels, LUTs, RTX Video HDR, and HDR wide-gamut restoration.
  • Anti-aliasing and beauty tools: FXAA, SMAA and MLAA, plus face-based beauty and restoration layers.
  • Export: an offline queue with MP4 or MKV output.

The author quotes performance costs on an RTX 5070: FXAA at 0.15 ms and SMAA or MLAA at 0.30 ms for standard 4K anti-aliasing fixtures, and a single-face mesh beauty effect at 0.93 ms for 1080p and 1.42 ms for 4K. The README says costs vary with content and hardware, and that no local AMD speedup percentage was measured.

Why the GTA6 and DLSS 5 tagline is a problem

The tagline combines three things: a console, a specific game, and a specific NVIDIA technology. The README body supports none of the three as a tested claim.

  1. No game test is shown. The text describes the PS5 source as Remote Play with an IP address, a PSN account ID and a pairing code. It does not publish a GTA6 capture, a frame-time log or a latency figure.
  2. DLSS 5 is hedged by the author. In the Limitations section the README says community runtimes are not vendor certification or a complete official DLSS 5 integration. For what NVIDIA itself has said about DLSS 5, see our coverage of the RTX Spark game developer update.
  3. Frame generation is not the same as a 120 FPS game. The README states that "submitted FPS is not physical display FPS or end-to-end latency." Generated frames add delay, and a Remote Play stream is capped by the console's own output and by your network.

We are not saying the tool cannot upscale a stream. We are saying that the headline number is a marketing line, and the author's own caveats are the better guide.

How do the sources work?

table · 2 cols
SourceSetup per the README
Local files and imagesDrop into the window
Capture cardSelect device, resolution, rate, pixel format and audio input; match PQ, HLG or 709 and limited or full range to the signal
PS5Enable Remote Play, discover or enter the IP, use the PSN account ID and the eight-digit pairing code
PCInstall and configure Sunshine yourself; Sunshine is not bundled; PIN-pair and connect
XboxDevice-code sign-in; the README calls it unofficial and experimental, with account, service and console restrictions
Screen capturePick a window or display; avoid capturing Veyra itself

For game mode, the README recommends a borderless game window or a monitor. It also describes a PS5 controller path through a capture card, where the console must be paired first.

What are the limits the author lists?

The README's Limitations section is candid, and it is the most useful part for deciding whether to try it:

  • RTSS overlay compatibility was verified locally only with RTSS 7.3.7 on an RTX 5070. GamePP and other overlays remain unverified.
  • Real-console Xbox long runs and console or VRR capture still need field retesting.
  • RTX 30 and 40 community NR and FG, FSR 4 ML hardware, some NR timeout and VRAM growth reports, and cross-device streaming have unresolved or unverified cases.
  • AMD NR requires driver HIP 7, and the AMD models are experimental.
  • Node mode in version 2.0.8 does not support offline export or List mode's global NR protection region.
  • Video export uses at most internal 1080p inference, so it is not native 4K NR.

What are the licensing and trust questions?

This is where a "GPL-3.0" badge needs reading carefully.

  • Original code is GPL-3.0. The combined Chiaki streaming application is also subject to AGPL-3.0 and an OpenSSL exception, with files under licenses/remoteplay.
  • Source excludes proprietary runtimes and models. The README says the source Git does not include them. The portable zips ship "approved runtimes", so what you run is more than what you can rebuild from source.
  • Release integrity. Manifests audit publisher files, but the README says they do not hash-lock user DLL replacements.
  • Individual-owned repo, large changelog. The repo shows 1,336 commits and releases from 2.1.1 on October 27 to 2.1.5 in early November. That is fast, which means fast fixes and fast regressions.
  • Donations. The README links Patreon and a US$5 coffee page and says donations do not restrict features.

How should you try it safely?

  1. Download only from the repository's GitHub Releases page, not from reposts or video descriptions.
  2. Check the published checksums and scan the archive.
  3. Extract into a new folder. The README warns not to overwrite a new package with an old veyra.ini or mixed DLLs.
  4. Start with effects off, confirm picture and sound, then enable one effect at a time.
  5. For PS5 or Xbox, treat sign-in codes and account IDs as credentials. Do not paste logs with private data into issues; the README itself says to review private data before sharing.
  6. Watch GPU temperature and VRAM, since layered NR and frame generation are heavy.

For readers who run local AI workloads on the same GPU, our posts on a GPU task manager for local LLMs and RTX Spark laptops hands-on help you budget VRAM across jobs.

How does it compare with cloud video upscalers?

Veyra runs on your GPU, in real time, and costs no per-minute fee. Cloud tools bill per second. For example, we tracked a 4K video upscale price cut claim that landed at a stated per-second rate. The trade-off is clear: cloud services handle generative upscaling of finished clips with no local hardware, while Veyra targets live playback and capture on an RTX or Radeon card you already own. They solve different jobs.

For the NVIDIA side, our look at the RTX Spark ecosystem launch covers the vendor's own tools.

What are people asking?

Is this an AI tool or just a shader pack? It uses neural models for SR and NR (RTX Video SR, vendor DLSS, FSR and community noise-reduction models), plus conventional shaders for anti-aliasing and colour. It is an enhancement pipeline, not a generative model.

Will it add input lag to game streaming? Yes, some. The README offers NVIDIA Low Latency modes and a low-queue option, and warns that submitted FPS is not end-to-end latency. Measure it yourself.

Does it work on a Steam Deck or Mac? No. The README states Windows 11 x64 and DirectX 12.

Why is it trending? Our read: the tagline, the DLSS 5 buzz, and a large reported download count. The README, not the tagline, is the evidence.

What this means for you

If you own an RTX card and watch low-resolution video or capture-card sources, Veyra is a serious, well-documented free option, and the changelog suggests active support. If you came for "GTA6 at 120 FPS," wait for tests you can reproduce. Either way, the right way to judge it is to run your own source with effects off, then on, and compare picture, latency and temperatures.

Related reading

  • NVIDIA RTX Spark game developer update and DLSS 5
  • RTX Spark ecosystem launch
  • RTX Spark laptops: a week of hands-on tests
  • A GPU task manager for local LLMs
  • Flux video upscale price claim, checked

Official source: the Veyra-NRVideo repository and README. For the vendor side of DLSS, see NVIDIA's DLSS page.

Star counts, versions and README statements are accurate as of November 13, 2026 and may change.

Spotted something out of date? Let us know.
Daily briefFrom Yash · 350K+ students taught

Catch up on AI

One AI story explained in plain English, plus one thing to try. Every day, free. Unsubscribe anytime.

  • ✓ Free
  • ✓ One email a day, no spam
  • ✓ Unsubscribe anytime
Yash Thakker

Written by

Yash Thakker

Yash is an AI expert with over 300K learners. Join his workshops →

View Yash Thakker in People in AI →

Related posts

Nov 16, 2026

MMFlash: Speculative Decoding for Vision-Language Models Explained

Text-only speculative decoding is mature; multimodal models stall when drafters guess image-conditioned tokens poorly. MMFlash from National University of Singapore measures visual dependence per token, upweights hard visual guesses, and sparsifies video attention—claiming up to 3.58× lossless speedups versus EAGLE-3 and DFlash baselines.

Nov 13, 2026

Chrome DevTools MCP: Let Your Coding Agent Debug a Live Chrome Browser

Chrome DevTools for agents, shipped as the chrome-devtools-mcp repo, lets a coding agent drive and inspect a live Chrome browser through MCP or a CLI. It trended on GitHub again in week 46 with 55,000 stars. Here is what it does, how to install it, and what to switch off.

Nov 13, 2026

LobeHub: Super-Agents in Your Team Chat, With 85K Stars and a Community License

LobeHub, the open-source chat and agent workspace, now pitches itself as "Super-Agents on Your Team". It trended on GitHub in week 46 with about 1,200 new stars in seven days and 85,000 in total. Here is what it does, how to self-host it, and what its license means.