OpenAim is a new browser-based FPS aim trainer with extensive customization options, colorblind-friendly themes, and data analysis features.
I played a lot of Valorant and got mad, so I made an aim trainer that analyzes your raw crosshair movement to explore your raw motor and perceptual weaknesses instead of scoring scenarios.
It also chooses sens and difficulty as part of the tasks, and makes playlists that are optimal difficulty for you to learn and progress faster!
# OpenAim
Source: [https://openaim.pramit.gg/](https://openaim.pramit.gg/)
### Appearance
ThemeTarget colorPresets are colorblind\-friendly against every theme\.Target stylePlayfield backgroundArena moodEarned backdrop palettes — free play only\. Locked ones show what unlocks them\.
Target finishEarned target skins — free play only\. Rated drills always use the standard target\.
Living sceneBackdrop shifts between drills & drifts slowly to fight visual fatigue\. Never touches targets or contrast\.Reduced eye strainDims & warms the scene and UI for long sessions — lowers luminance load\. Targets keep full contrast\.MotionReduce disables UI transitions; Auto follows your OS\.
### Crosshair
StyleSizeThicknessGapColorthemeOutlineKeeps the crosshair visible over targets\.On\-target reactionOff by default for clean telemetry\.
### Feedback
Hit markerHit soundVolumeMiss flashRed flick at the crosshair on a missed shot\.Shot feedback soundsMissed shots and magazine reloads\.Blocked\-shot cueMulti\-tap drills cap the fire rate\. Marks pulls that came too early to fire, so they read as your timing, not lag\.Tracking toneSustained lock tone that climbs while your crosshair holds a target\.Off\-screen spawn cuePanned alert pointing to targets that spawn out of view\.Off\-screen spawn markerA brief edge chevron toward a target that spawns out of view — the visual twin of the spawn cue\.Pace metronomeCoached drills: one beat per pace budget\.Reward momentsConfetti \+ a flourish when you set a record, rank up, or hit a streak\.Eye\-rest remindersA gentle 20\-20\-20 nudge between drills on long sessions\.HUD elementstimerscoreaccfpsZen modeHide the whole HUD during runs; results still show everything\.
### Input & camera
Sensitivity° per mouse count\.Mouse DPIOptional\. Enables cm/360 and hand\-travel modeling\.Import sensConverts your in\-game sensitivity\.Field of viewHorizontal; stamped into every replay\.Invert YRaw inputWhether the browser delivers unaccelerated mouse motion\. OpenAim never accelerates; this is the browser/OS layer\.
### Coach
Challenge pointTarget success rate drills are solved to\.ExplanationsFull detail always stays behind the ? chip\.Session lengthDrills per coached session\.FocusLet the coach choose, or lock the session to one axis\.AutoGainSens auto\-tuning from measured under/overshoot\.
### Data
Capture levelSummary skips raw mouse/camera streams; deep analysis needs Full\.Auto\-download replaySave the \.oar after every completed run\.Your data
### Aim Commons
Anonymous drill summaries are contributed by default to train the shared, science\-based player model everyone starts from and that renormalizes your ratings to real peers\. No raw mouse traces, no identity, no device info — and you can opt out or revoke below anytime\. Raw replays and the public leaderboard stay off unless you turn them on\.
Contribute runsSends drill summaries \+ capability features \(no raw mouse, no device info\) to sharpen the population model everyone starts from\.LeaderboardRank on the task\-Elo boards under a public handle\.HandleYour public leaderboard name\.Share replaysAdvanced: uploads raw mouse\-movement traces \(mildly biometric\) for deeper analysis and spectating\. Requires the two above\.Statuslocal only
Settings that change the meaning of telemetry \(FOV, sens, target style, crosshair, invert, capture level\) are recorded in each replay's \.oar header, so analysis reflects the settings the run used\.
Homegames is a free, open-source game platform that lets users play, make, and share games in the browser, with a code editor and live multiplayer preview. It has been developed as a side project since 2018.
An AI agent using a 30-billion-parameter model leverages headless screenshot loops to autonomously complete a raytraced first-person shooter demo written in C, showcasing advanced autonomous coding and debugging capabilities.
OpenEnv, a tool for creating agentic execution environments like terminals and browsers, is transitioning to a more open governance model with a committee including Hugging Face, Meta-PyTorch, Nvidia, and others to promote open-source agent training.