Your device.
Your first live session.

Android setup — no laptop required

Download Android APK

The signed APK supports Android 7 or later on 64-bit ARM and x86 devices with 4 KB memory pages. Devices using 16 KB memory pages are not supported by this release. This is a direct download, not a Google Play listing; the Windows installer does not install on Android.

1. Download the APK from this site and check the SHA-256 on the download page. Open it; if Android asks, grant installation permission to the browser or file manager you used, then turn that permission back off after installing. Keep Play Protect enabled. Do not bypass an incompatible-device or security-policy block.
2. Connect the Android device and your PS5 or gaming PC to the same private home network. Open Your Pit Box on the tablet; the laptop is not required for the native Android engine.
3. Keep setup step 2 open. Enter its UDP IP Address and port in the game; set telemetry On, broadcast Off and format 2026. The usual port is 20777; use the value shown. Connection also has these details.
4. Start an on-track session and confirm Receiving telemetry. A listening socket or a saved lap chart alone does not prove that fresh packets are arriving.
5. For radio, save and test the provider key in Connection, allow Android microphone access, and enable voice in the app. Speech services need an OpenAI key and internet access; the dashboard can receive telemetry without voice. The optional billing and key creation instructions apply to either app.
6. Check microphone input separately from the reply. If the app hears input but you hear nothing, check the selected output, media volume and Bluetooth routing. A key test does not verify the microphone or speaker.

Future updates to the release app use the same permanent signing identity. Install a verified release update over the release app; do not remove its saved data.

Before a full-length race, check your microphone, selected speaker or headset, wake phrase and background settings on the device you will use. A short radio check verifies that setup, not every Bluetooth device or power-saving mode.

Bluetooth stability on Galaxy Tab S11 Ultra (Android 16): a combined Wi-Fi/Bluetooth disconnect was reproduced with Bluetooth audio connected during a 60 Hz telemetry test. The cause is under investigation; version 5.3.2 does not yet resolve this observed condition. A successful test using the tablet speakers does not verify your Bluetooth headset.

Forward live telemetry to another app or device

Optional UDP forwarding is separate from saved-session transfers and from receiving game telemetry directly. A forwarded packet is best-effort, not a delivery receipt.

1. First confirm that the sending engineer shows Receiving telemetry from the game.
2. In Connection → UDP forwarding, add the destination app's address and listening port. Prefer a trusted destination on the same private network.
3. Enable that target and check the receiving app for fresh telemetry. The sender's packet count alone does not prove delivery.
4. Avoid forwarding loops or sending both a direct and forwarded feed to the same receiver. If reception is unreliable, disable forwarding and point the game directly at the engineer you are using.

Beta forwarding does not copy your saved library, API keys or microphone permissions. Do not forward telemetry to a public address unless you intend to share it with that destination.

Compare the laps you choose

Open Setup Lab → Open Test Engineer, or Analysis → Test Engineer. Choose the current session or a saved session. Automatic runs remain available alongside your own named lap groups.

Define Stint A, B and C

1. Open Define my lap groups. Use Add group and give it a name, then choose From lap and Through lap and select Use this range. Check or uncheck individual laps to refine the selection.
2. For a starting point, set Number of suggested groups to 2–12 and choose Suggest groups. You need at least one recorded lap per group. Suggestions follow session order and favour recorded tyre, setup and session changes; they do not cluster unrelated fast and slow laps.
3. Review every group, then choose Save groups. A lap can belong to one custom group. Laps left unassigned, including newly completed laps, stay in the recording and can be added later. After a rewind, the timeline label distinguishes repeated lap numbers.
4. Under Choose from, select My saved lap groups. Choose Baseline A and Test B, then Stint comparison — any compound and Compare stints. Choose Automatic runs to return to event-based runs.

Use stint comparison for medium versus hard tyres or other experiments you want to review. It describes selected clean pace and sectors alongside the conditions; it does not isolate a tyre or setup effect. A group containing different tyre sets, setups or timelines has no single pace-trend estimate. Sector differences use laps with all three sectors recorded, which may be a smaller cohort than the overall pace comparison.

Keep the explanation with the lap

Open Lap context and driver reports, choose the affected lap range and reason, describe what happened, then select Save lap note. Check Exclude these laps from pace comparisons only when that assessment applies to every selected lap. Leave it unchecked for context such as a handling observation. Use Edit note or Remove note if the assessment changes; removing a note also removes any exclusion supplied by that note.

With a reasoning provider configured, you can tell the engineer, “Lap 6 was compromised because I was held up in sector two,” or ask it to note an issue on a named group. Wait for the engineer to confirm that the note was saved. A report about “this lap” stays attached to the lap where the request started, even if you cross the timing line during the reply. An unfinished lap's note remains pending until that attempt is recorded.

Later reviews, comparisons, pit-entry debriefs and exports include the saved context. Driver reports and engineer interpretations are labelled separately from measured air/track temperatures, fuel, tyre age and traffic evidence. A proximity flag shows nearby cars; it does not measure how much time traffic cost. Missing traffic evidence does not establish clear air. Notes never replace the original telemetry.

Understand a slow lap in the middle

A slow lap stays in the same run. With stable measured conditions and consecutive neighbouring laps back at baseline, a slowdown greater than 2.5 seconds or 3% (whichever is larger) is labelled isolated slow lap and excluded from clean pace. The neighbours must agree within one second or 1% (whichever is larger). This identifies a temporary pace change; it does not establish whether you cooled the tyres, made a mistake or met traffic. Missing context and sustained changes do not establish an isolated slow lap. Excluded laps and their reasons remain visible.

Run a matched setup test

A and B can come from one session or separate saved sessions at the same circuit. Choose B’s session, then choose its automatic run or custom lap group. Each session keeps its own notes and exclusions. Groups, notes and session exports below the comparison belong to A; open B’s session to edit its saved notes. Check the comparison limits: matching laps does not verify the same game version, car, formula or car performance.

1. Use Test plan and conclusions to record your objective for the whole session, an automatic run or a saved group.
2. Complete at least three clean timed laps. Returning to the pits creates a debrief on Drive; it can also be spoken when proactive radio is enabled and the engineer is free. Leaving the pits starts the next run.
3. Change one setup setting. Keep compound and starting fuel comparable, then repeat the run in clear air. Runs also split when the recorded setup or tyre set changes.
4. Select A and B and choose Matched setup test, then Compare matched laps. Each group must contain one recorded setup. Matching requires the same compound and weather, tyre age within one lap, fuel within 3 kg and track/air temperatures within 3°C. Invalid, pit, flagged, traffic, reported exclusions, outlier and incomplete laps are excluded. At least three matched pairs are needed for a verdict.
5. Read the B-minus-A lap and sector differences: negative means B was quicker. Save your conclusion, then choose Export session report for text or Export JSON. On Android, choose a destination in the file picker.

Session Review → Review runs / export report opens a saved session's Test Engineer report. Cross-session setup tests also require the same recorded session mode; known differences in circuit length prevent comparison. Older sessions remain available, but missing setup or condition evidence cannot be reconstructed. Observed pace trend includes fuel and driving effects; a matched comparison does not prove the setup caused the difference.

For a race, Strategy → Who am I racing? ranks the closest projected finish gaps after estimated remaining stops. Tyre age, current position and confidence are shown. Unknown gaps and predicted stops remain estimates.

Choose a baseline, then test it

Open Setup Lab, choose your circuit and conditions, then select Stable or More rotation. Both references cover all 24 calendar circuits for F1 25: 2026 Season Pack. Choose Race, Quali or Mixed; qualifying uses separate published values where supplied, while Mixed uses race values. Settings are grouped like the game garage. Open Sources & setup details for the original references, review date and adaptations.

Circuit baseline preserves the selected source values. Choose Personalise for me to apply your preferences and applicable live feedback: Small adjustments makes small changes to your current car, Moderate changes blends toward the reference, and Start fresh starts from it. Personal changes are suggestions to test, not proof of a pace gain. Old aggregate setup scores and universal tyre-temperature rules no longer alter the reference.

References target dry conditions. Wet or unsupported circuit/style selections report when no applicable reference is available. For qualifying under parc ferme, check the source's race-wing and brake-pressure guidance before committing. Published setups still need testing with your car, assists and driving style; use Test this setup to open Test Engineer and compare matched runs.

Enter the settings in the game garage; YPB does not apply them to the game. During a race, only a separately identified permitted front-wing adjustment is offered for a pit stop. Unsupported numerical lap-time gains and tyre-life multipliers are no longer presented as setup predictions.

For Android connection interruptions, the Connection page now reports OS-observed network losses and severe thermal status. Those readings help distinguish a device/network interruption from missing game telemetry.

If a check does not pass