This Bluetooth latency test measures the end-to-end audio delay of whatever you are listening through: phone, computer, TV link, car stereo or Bluetooth headphones. Put the device on, tap in time with a steady beat for about ten to fifteen seconds, and the tool reports your chain's estimated delay in milliseconds. It runs entirely in your browser, with no recording and no upload. The number is an estimate of your whole output chain, not a codec specification; published reference figures sit just below the tool.
Audio could not start in this browser, so the beeps cannot play. Try a current version of Chrome, Edge, Firefox or Safari.
Put on the device you are testing and turn its volume up. Tap the pad (or press Space) exactly when you hear each beep, like following a beat. There is no visual metronome on purpose: follow the sound, not the screen.
- Runs entirely in your browser. Nothing is recorded or uploaded.
- The result is an estimate of your whole audio chain, not a lab measurement. See the method below.
Reference figures on this page were checked October 2026. The tool above runs locally in your browser: nothing you hear or tap is recorded or uploaded.
How this Bluetooth latency test works
The tester plays a steady beep at a fixed 500 ms interval through the device you want to measure. You tap along with what you hear, the way you would nod along to a song. Each tap is compared against the scheduled beep time and folded into a plus or minus 250 ms window; the tool drops the first four taps while you lock onto the rhythm, then reports the median of the next twenty together with their spread.
Tapping along, rather than reacting, is the core of the method. A reaction carries the time your own response system needs to answer a surprise, which dwarfs the delays being measured; anticipating a steady beat cancels that constant lag, leaving the offsets settled near the true delay of the chain.
Expect some tap-to-tap scatter even when you are locked in. The tool reports the middle 50% spread so you can judge the quality of each run, and a retest takes seconds. Taps that settle slightly early and taps that settle late are read the same way: the size of the offset from the beat, not its direction, is the estimate.
What your number means
The estimate covers the whole output chain: the browser audio stack, the operating system, codec negotiation, the radio link and the conversion inside your headphones or receiver. It is not a property of any single codec. To put your result in context, here are the published, independently measured reference figures; none of these numbers is ours.
| Chain or codec class | Published figure | Source / as of |
|---|---|---|
| Wired USB | 13 ms | RTINGS bench, 2026-09-22 |
| Wireless 2.4 GHz dongle | 23 ms; gaming dongle class spread of 22 to 42 ms | RTINGS bench, 2026-09-22 |
| aptX Adaptive, Low Latency mode | Under 80 ms, a bench verdict | RTINGS bench, 2026-09-22 |
| aptX Low Latency, official design target | Approximately 40 ms | Qualcomm, 2026-09-22 |
| SBC | 259.8 ms truly wireless average; 308 ms four-phone average | RTINGS and SoundGuys, 2026-09-22 |
| aptX | 316 ms four-phone average | SoundGuys, 2026-09-22 |
| LDAC | 324 ms four-phone average | SoundGuys, 2026-09-22 |
| AAC | 369 ms four-phone average | SoundGuys, 2026-09-22 |
The full compilation, with rigs, sample sizes and caveats, lives on the measured codec latency bench. The same codecs swung 244 to 484 ms between phones in the SoundGuys test, so treat these classes as neighborhoods rather than promises. If your result sits far above every row, the usual suspects are the source device, an app adding its own buffering, or a crowded 2.4 GHz band; the Bluetooth audio delay guide walks through the fixes.
What shifts your number
- The codec is set by both ends. A chain negotiates one codec that the source and the receiver both support; if either end lacks a low latency option, the chain falls back. The aptX LL profile documents the approximately 40 ms design target and what it is anchored to.
- Processing on the source. Operating system effects, equalizers and some apps add buffering before audio ever leaves the phone or computer. Running the test from two different sources on the same headphones often moves the number more than the codec does.
- The radio environment. Distance, walls and a crowded 2.4 GHz band cause retransmissions, which show up as both delay and spread. A second run in the same room tells you how stable your setup is.
FAQ
It is an estimate, not a lab measurement. The spread reported alongside the median shows how tight your taps were, so judge each run by its own spread. The test suits comparing two chains, or checking whether a fix actually helped, rather than quoting a specification.
Reacting to a sound carries the time your own nervous system needs to respond to a surprise, which is far larger than the delays being measured. Anticipating a steady beat cancels that constant lag, so the tap offsets settle near the actual delay of the audio chain. This is also why the page shows no flashing metronome: a visual cue would measure your eyes instead of your ears.
Everything between the browser scheduling a beep and the sound reaching your ears: the browser audio stack, the operating system, codec encoding, the Bluetooth radio link and the conversion in the receiving device. It is an end-to-end figure, so it normally sits above any single codec's published figure.
Yes, as long as the beeps travel through the chain you want to measure. With headphones or a speaker, simply listen through it. For a car or a TV, run this page on the source device of that chain: a phone paired to the car stereo, or the TV's own browser with its Bluetooth output connected.
blafili B3 Bluetooth Receiver
XLR, RCA, optical and coax outputs with an OLED front panel, built on the QCC5125 + ES9018K2M platform. A wired amp or receiver, put on a wireless source.
- RTINGS, "Test Bench 1.6 - Latency: R&D Article": wired USB 13 ms and wireless dongle 23 ms (Audeze Maxwell Wireless); gaming dongle class spread of 22 to 42 ms; "<80 ms is appreciably low" (aptX Adaptive Low Latency); SBC averages of 170.3 ms (gaming headsets) and 259.8 ms (five truly wireless earbud products). rtings.com/headphones/learn/research/latency, accessed 2026-09-22, quoted from our latency bench.
- SoundGuys, "Android's Bluetooth latency needs a serious overhaul": four Android phones feeding a receiver and instrumentation, averaging SBC 308 ms, aptX 316 ms, LDAC 324 ms, AAC 369 ms; phone-to-phone swing of 244 to 484 ms. soundguys.com/android-bluetooth-latency-22732/, accessed 2026-09-22, quoted from our latency bench.
- Qualcomm aptX Low Latency official page: latency of approximately 40ms. aptx.com/aptx-low-latency, accessed 2026-09-22.