Measure your FPS
Run a fair before-and-after test and understand the result.
On this page
Where: the CS2 tab.
A benchmark is a repeatable performance test. You do not need one to play, but it helps you check whether a change helped.
| Your question | Choose |
|---|---|
| Did TinyBoost help? | CS2 Automated Tweaked FPS Benchmark |
| Did my own change help? | Manual Benchmark |
Both use the same Dust 2 benchmark map and compare BEFORE with AFTER.
Before you start#
Save your work, close CS2 and sign in to Steam. Finish any active TinyBoost session, stress test or reset. Allow about 10 to 15 minutes, depending on your chosen settings, and leave the PC alone during each run.
The first time, follow these prompts:
- Read the Intel PresentMon notice and click Continue & Download PresentMon. PresentMon is the tool that records frames.
- Subscribe to the benchmark map on the Steam Workshop page that opens. Wait for it to download, then click I've Subscribed, Start Benchmark.
- In Unlock Steam, unlock Steam Family View if you use it. Click Steam family is OFF. Start the Benchmark when ready.
During a run, do not use the mouse, keyboard or Alt+Tab. Keep CS2 in front.
CS2 Automated Tweaked FPS Benchmark#
- Click Run Again BETA on the CS2 FPS Benchmark card.
- Choose Duration per run: 20 to 104 seconds. Click START and follow the preparation prompts.
- Wait for the BEFORE run.
- At Initial Benchmark Complete, click Continue to proceed, or Cancel Benchmark to stop.
- Leave the PC alone for the three AFTER runs and cleanup.
What happens#
TinyBoost first records CS2 with its optimizations off. It then applies its optimizations and records the same map three times. A brief black screen or game relaunch can happen between stages.
Your CS2 video settings stay the same for both sides. The chosen run duration is also the same.
When finished, TinyBoost turns its optimizations off and opens the report in your browser. Wait for cleanup and read any restoration messages. Press START later if you want to play with TinyBoost active.
Manual Benchmark (your own changes)#
Use this for a graphics setting or another change you can make within the same session. Restarting Windows can interrupt the comparison.
- Click Start Manual Benchmark.
- Choose 1 to 5 runs for each side and a Duration per run of 20 to 104 seconds.
- Fill in BEFORE name and AFTER name, such as “Old settings” and “New settings”.
- Click Start Manual Benchmark and follow the preparation prompts.
- When BEFORE runs complete appears, make one change.
- Click Start the AFTER benchmarks and leave the PC alone.
TinyBoost records the game but applies no optimization tweaks. If you choose several BEFORE runs, their average becomes the baseline.
To undo your change: restore it yourself after the test. This mode does not undo settings you changed by hand.
Reading the results#
Start with these three numbers:
| Number | What to look for |
|---|---|
| Average FPS | Overall frame rate. Higher is faster. |
| 1% Low Average | Performance during slower moments. Higher is better. |
| Stuttering Time | Time spent in unusually slow frames. Lower is better. |
A higher average with worse lows can still feel less smooth. Small differences can be normal variation. Compare all runs and repeat surprising results before keeping a change.
The report also includes more detailed readings, such as P1 FPS, 0.1% Low Average and PC Latency. You can leave these aside at first. A missing reading does not mean the whole test failed.
Past results#
Click Results History, then Open Report. Previous Benchmark Results or Show Previous Results reopens your last result from the benchmark card.
If a run fails#
Read the message first. Check that Steam is signed in, the map is downloaded, CS2 stayed in front and no other TinyBoost job is running.
To stop, use Cancel and wait for cleanup. If the PC restarted during a test, reopen TinyBoost so it can run its recovery steps. Read any restoration messages.
If it fails again, send support the exact message and say whether it happened during BEFORE or AFTER. See what to send.