Android + OpenCL implementation of Pocket-TTS — on-device on non-flagship phones (Adreno 6xx)

Hi! Sharing this as an issue since the HF community tab for kyutai/pocket-tts is disabled — maintainers, feel free to close/convert this; just wanted it to be findable for Android users.

I wanted to share an Android + OpenCL implementation of Pocket-TTS, in case anyone wants to run it on a phone:

• Try it: Edgi on Google Play — runs fully on-device, no cloud.
• Open source: the app is built on top of the open-source adreno-llms inference engine — https://github.com/a8nova/adreno-llms — pure C++/OpenCL with hand-written kernels tuned for Adreno, covering the FlowLM backbone, flow-matching denoiser, and Mimi SEANet decoder.

It's tuned and tested on Adreno 6xx GPUs — the GPU class in mid-range and older Android phones (verified on a 2020 Motorola Razr / Adreno 620) — and should run on most arm64 Android phones with OpenCL, though the optimizations are Adreno-specific. On the Adreno 620 it synthesizes 24 kHz speech at RTF ~1.47 with 8 selectable v1 voices, bit-exact vs the fp32 reference.

Hope it's useful — happy to answer questions!