Valorant's Vanguard runs at kernel level 24/7 — even when you're not playing. That makes internal cheats nearly impossible. External ESP, however, reads from outside the process and Vanguard has no reliable way to detect it.
Recommended ESP Configuration
| Setting | Value | Why |
|---|---|---|
| Box type | 2D box, thin line | Less visible if screenshotted |
| Box opacity | 40–55% | Looks like UI clutter |
| Name tags | OFF in ranked | Name tag ESP is the most reviewed |
| Distance | OFF | Numbers are obvious in demos |
| Health bar | ON — adds legitimacy | Can explain pre-fires |
| Skeleton | OFF | Only useful for aimbot — too obvious |
| Visible check | ON | Only show enemies you "could" see |
Stream-Proof Mode
Enable stream-proof (the cheat renders in a window that's invisible to screen capture APIs). Riot's manual review team watches streams and VODs — if your ESP shows on stream, you're reported immediately.
Aimbot — Keep It Subtle
Vanguard doesn't detect aim patterns directly, but Riot's manual ban team does. Stick to:
- FOV: 10–15° max
- Smoothing: 50+ (slower = more human)
- Trigger delay: 80–120ms after crosshair on target
- Never activate on pistol rounds — too easy to spot in demos
Free Download
The Cosmo Cheatz Valorant build is external, free, and updated within 24 h of Vanguard patches. Check status before every session.
Screen Capture Detection Across OBS, Medal, Nvidia
Stream-proof rendering is only as good as the capture path it defeats. OBS with Game Capture (the modern Windows 10/11 hook) will pick up overlays drawn in a separate layered window more often than OBS Display Capture, which composites the desktop and catches almost everything. Medal and Outplayed both hook DXGI Desktop Duplication, so a cheat claiming stream-proof but not masking that API will still appear in clips. Nvidia ShadowPlay uses NVFBC on older drivers and the newer NvCapture SDK on 550-series and above, which behaves differently again. Before trusting a build, record a ten-second clip in each tool you actually use and play it back on a second monitor. Look specifically for ESP boxes during a smoke or flash, where the overlay has no natural background to hide against. Setting the cheat's render target to a topmost transparent window defeats Display Capture but not Game Capture in windowed borderless. Fullscreen exclusive is the safer mode for masking, at the cost of alt-tab speed.
What Changed Between Vanguard Builds
Vanguard's protection level has drifted over the past two years. Builds shipping after the mid-2025 rollout of the TPM 2.0 and Secure Boot requirement on Windows 11 tightened the hypervisor checks, and several external readers that relied on physical memory handles stopped working until they switched to a different driver-backed read path. On Windows 10 the same builds remained largely functional because the TPM enforcement was never backported. A second change arrived with the introduction of periodic integrity scans during a match rather than only at launch, meaning a reader that worked at the main menu could fail midway through a competitive game. The failure mode is usually a Vanguard error 57 or 68 followed by a restart demand, not an immediate ban. Verifying a build works means loading into a deathmatch, playing four or five minutes, then alt-tabbing and checking that the overlay still reads enemy positions after the scan window closes.
Timing ESP Refresh Against Server Tick
Valorant servers tick at 128 Hz, and your client renders at whatever your monitor and frame rate allow. A cheat that redraws positions every frame from stale memory will drift behind where an enemy actually is, producing boxes that trail a sprinting Jett or lead a strafing Reyna. Most competent external readers poll the entity list at 60 to 120 Hz and interpolate between the last two samples, but a badly configured refresh interval below 30 Hz makes boxes visibly lag on wide swings. The practical test is a custom game with a friend strafing left and right at full speed while you watch the box against their hitbox on screen. If the box lags more than roughly one frame at 144 Hz, the poll rate is too low or interpolation is off. Setting the refresh to match your monitor is wasted work above 144 Hz and raises CPU overhead, which some players notice as input delay on a mid-range CPU running the game at 300-plus fps.