You hover over the Buy button and feel the old, familiar gamble—will this run, or will it stutter into a refund? I’ve clicked purchase on promises of smooth 60 FPS and been burned by 20. Now Valve may hand you a number before you risk your wallet.
On Steam’s storefront, dataminers found a new FPS report
SteamDB and a couple of eagle-eyed tweeters pulled a string and found the code: Valve is building a feature that estimates expected frames per second based on other Steam users’ data. Lambda Generation and Roadrunner were among the first to flag the discovery, and the snippet even includes a short prompt: “Select an App and a PC config to get a chart of estimated framerates, based on the framerates of other Steam users.”
Steam will soon tell you how much FPS you may get according to your PC specs by taking other Steam users with similar hardware.Found by Roadrunner on SteamDB. pic.twitter.com/H0opyA79Eo
— LambdaGeneration (@LambdaGen) April 4, 2026
The feature looks manual: you can pick hardware profiles, save settings, and generate a chart. If it lands in Steam’s beta or main branch soon, you could compare your rig—Intel, AMD, or NVIDIA GPU—against a pool of real-world players before buying.
Can Steam predict FPS before I buy?
Yes, but not like a crystal ball. Valve’s proposed system uses crowdsourced telemetry—stats players already attach to reviews and reports—to estimate framerate for users with similar specs. That data-driven approach is promising: it’s less guesswork than forum advice and more empirical than vendor benchmarks.
On review pages, hardware tags already point the finger at performance
You’ve probably seen the tiny spec tags in reviews: CPU, GPU, RAM, driver versions. That feature gave players a voice; this FPS estimator could give that voice numbers. If the stats are solid, it acts like a weather forecast for your GPU—short, practical, and oriented around what actually happens to people who have the same machine as you.
This could cut down on angry one-star reviews from players who didn’t match developer minimums. It could also raise expectations: a promising FPS number might make buyers angrier when a day-one patch or driver quirk changes the reality, shifting blame toward developers and Valve alike.
How accurate will Steam’s FPS estimates be?
Accuracy depends on data quality and sample sizes. Big titles with thousands of telemetry points will produce tighter estimates; small indies with ten reported runs won’t. Drivers, OS updates, background apps, overlays, mods, and custom settings all add variance. I’d treat the number as a strong hint, not a guarantee.
At checkout, your expectations and refunds are already in tension
When you decide whether to buy, you balance hype, reviews, and refund windows. Valve’s estimator shifts that balance by adding a measurable expectation. For developers, it’s both risk and relief: numbers may quiet speculation when performance is solid, or spotlight problems faster when it’s not.
Think of it like a scoreboard that whispers the truth—sometimes encouraging, sometimes brutal. The rollout will matter: whether Valve arms players with per-resolution, per-graphics-preset charts, or a blunt average, will shape how players interpret the data.
On rollout and responsibility, the small print will matter
SteamDB shows this change in code; Valve could push it to beta first. If it uses the same telemetry that powers hardware attachments in reviews, it should improve over time as more players contribute data. But privacy, sampling bias, and regional hardware skew are real considerations—some GPUs and drivers are far more common in different markets.
Platforms and partners will notice. NVIDIA, AMD, and Intel drivers shape performance curves; anti-cheat and overlays can alter them. Developers will have to clarify supported settings, and players will likely start referencing estimator charts in threads and streams.
I won’t pretend this solves every performance debate, but it turns guesswork into a conversation. Will seeing a predicted FPS before you buy make you happier or just angrier when your rig behaves differently?