Bugfix release (VAJ edition, io.vaj.tl). Install alongside stock Termux; pulls packages from the VAJ APT repo at first run.
Fixed
- Rate-limited API responses (HTTP 429) now include
Retry-AfterandRateLimit-*headers so OpenAI/Ollama clients can back off correctly instead of guessing. - Attempting to load an embedding model into the generation runtime now returns a clear error; embedding models are served on demand through the embeddings endpoints and no longer need to be loaded.
Pinned package state
- Bootstrap SHA-256:
88985cead42037536291fc2575dc04e22e4bceb1b4f1c6601e3c008ed09dc204 - APT repo:
https://repo.pathayam.xyz stable main - Repo signing key:
98C1464FEE61FCBA4E7F54AE86B650B212C8DB81
arm64-v8a debug APK only (no embedded bootstrap, ~54 MB). Verify with the attached sha256sums.txt.