Translate anything.

LYR - On-screen translation for Android.

Works with your favorite apps.

Instagram
TikTok
X
Prime Video
Weverse

Coming Soon

Understand everything.

Manga

Replaces what is inside the balloons, without breaking the panel flow.

Live translation

Subtitles keep up with the speaker while you watch.

Audio

Turns the sound playing on your device into subtitles.

Web

Any site — understandable as you scroll.

Research

Where does the speed come from?

Latency, OCR, quantization, self-hosting LLMs. The record of what I measured and got wrong while building LYR, on my own.

#032 A general-purpose model is too wasteful a vessel for subtitles Even a raw 8B with no compression at all left 15 critical mistranslations out of 97. The 27B, more than three times bigger, had 4 (p=0.015). Varying the strength of the compression doesn't move that gap. So the next suspect is the vessel itself — subtitles are short, constrained, bounded-context, and yet we use, whole, a vessel designed for free generation of arbitrary length. What I expect from specialization isn't only speed: emotional register, and accuracy from translating while looking at the video. Here is the design hypothesis, and honestly, what would falsify it. 22 min
#031 Five rebuilds and it still never beat chance — the cause was the labels, not the model Change the inputs, change the architecture, rebuild it five times, and accuracy still doesn't beat chance. The cause wasn't the model — it was ground-truth labels that were off by about a second. And broken labels cost you twice — they don't just produce a bad model, they had also been undervaluing a good implementation I already had as '25% misses'. 8 min
#030 Generation speed is decided by the byte count of the weights, not the parameter count Measured on the same GPU, 4B-INT8 and 8B-INT4 — a 2x difference in parameters — ran at almost the same speed (5.98ms against 6.50ms per token). Every token generated reads out all of the weights, so what decides speed isn't the amount of compute or the parameter count but the byte count. Step outside the premises, though, and the law breaks; get the measurement method wrong and a 2.35x gap disappears, or inverts. 8 min
#029 False fires on UI elements: 44% → 22%, without adding a single outer if A gate that's supposed to be watching for subtitles fires on a video app's UI controls and a browser's URL bar. Add a filter on the outside and it goes away. It did go away, and power dropped too. But that was a contract to keep buying insurance, every single frame, against the model being weak. How I fixed the label definition and made the model side solve it — and the one condition under which outer logic is still the right answer. 10 min

Read all 33

Now you can understand anything.

LYR - On-screen translation for Android.

Manga · Live streams · Audio · Web.

Google Play Coming Soon

I'll announce the release on social media.