Auriel Solaris
Get the apps
← The suite
NekoChat icon

AI chat

NekoChat

Private AI chat that runs on your phone.

Runs GPT-2 and Qwen3 models on the phone with its own C++ engine, using Vulkan, OpenGL ES or the CPU. It needs no ML framework, and your chats never leave the device.

Get NekoChat Source on GitHub
Version
0.1.3
Licence
GPL-3.0
Runs on
Android 5.0+
Origin
Original to the suite

See it in action.

  • Chats
    Chats
  • Qwen3-0.6B, streaming
    Qwen3-0.6B, streaming
  • Download from Hugging Face
    Download from Hugging Face
  • Weight precision
    Weight precision
  • Rose and Ocean themes
    Rose and Ocean themes
  • Read aloud with Kokoro
    Read aloud with Kokoro

What makes it different.

  • Uses the GPU when it can

    Auto mode tries Vulkan, then OpenGL ES 3.1, then the CPU. A GPU is used only after its results match the CPU in a self-test.

  • FP16, FP8 or FP4

    Weights are quantized as the model loads. FP8 halves memory with nearly identical answers, and FP4 quarters it.

  • A voice that stays on the phone

    Kokoro-82M reads replies aloud on the CPU, with 28 English voices. It works on de-Googled phones with no text-to-speech engine.

  • Bring your own model

    Download from Hugging Face with aria2, or pick a folder of .safetensors or .pt files. NekoChat checks config.json before it downloads anything.

Built with.

  • Kotlin
  • Jetpack Compose
  • C++17 / JNI
  • Vulkan · GLES 3.1 · NEON