Starstruk Logo MarkStarstruk

v4.4.3

danielmiessler/Fabricv4.4.3Jun 13, 2026by mudler

AI Summary

A maintenance release introducing new TTS backends (OmniVoice, Kokoro), Indonesian language support, and various dependency updates.

Key Highlights

  • New OmniVoice TTS backend with streaming and voice cloning capabilities.
  • New Kokoro TTS backend adding multilingual Piper voices.
  • Indonesian language support added to the React UI.
  • Locate-Anything-CPP backend for object detection.
  • Distributed mode enhancements with declarative per-model scheduling.

New Features

  • OmniVoice TTS backend
  • Kokoro TTS backend
  • Indonesian language support
  • locate-anything-cpp (object detection)
  • distributed declarative per-model scheduling

Full Release Notes

<!-- Release notes generated using configuration in .github/release.yml at master -->

## What's Changed
### Other Changes
* chore: :arrow_up: Update CrispStrobe/CrispASR to `d745bda4386ae0f9d1d2f23fff8ec95d76428221` by @localai-bot in https://github.com/mudler/LocalAI/pull/10260
* docs: :arrow_up: update docs version mudler/LocalAI by @localai-bot in https://github.com/mudler/LocalAI/pull/10259
* chore: :arrow_up: Update antirez/ds4 to `d881f2a05e8ff6bec001315a36b794b4aa310173` by @localai-bot in https://github.com/mudler/LocalAI/pull/10262
* chore: :arrow_up: Update mudler/parakeet.cpp to `9db92be63179a27201d3b88d5d40c545b2ac48ae` by @localai-bot in https://github.com/mudler/LocalAI/pull/10263
* feat(react-ui): add Indonesian language support by @dedyf5 in https://github.com/mudler/LocalAI/pull/10266
* chore: :arrow_up: Update ggml-org/llama.cpp to `4c6595503fe45d5a39f88d194e270f64c7424677` by @localai-bot in https://github.com/mudler/LocalAI/pull/10261
* feat(backend): locate-anything-cpp (open-vocabulary object detection via ggml) by @localai-bot in https://github.com/mudler/LocalAI/pull/10264
* fix(router): production-ready request router + auto-size batch for embedding/rerank by @richiejp in https://github.com/mudler/LocalAI/pull/10104
* chore(model gallery): :robot: add 1 new models via gallery agent by @localai-bot in https://github.com/mudler/LocalAI/pull/10270
* feat(parakeet-cpp): enable GGML_CUDA_GRAPHS in the cublas build by @localai-bot in https://github.com/mudler/LocalAI/pull/10273
* fix(darwin): publish sherpa-onnx and speaker-recognition images for darwin/arm64 by @localai-bot in https://github.com/mudler/LocalAI/pull/10275
* fix(crispasr): write piper TTS WAV at the model's native sample rate by @localai-bot in https://github.com/mudler/LocalAI/pull/10277
* feat(crispasr): bundle espeak-ng and add piper TTS voices to the gallery by @localai-bot in https://github.com/mudler/LocalAI/pull/10283
* chore: :arrow_up: Update mudler/parakeet.cpp to `b8012f11e5269126eddb7f4fd02f891a2ccc29b0` by @localai-bot in https://github.com/mudler/LocalAI/pull/10281
* docs: :arrow_up: update docs version mudler/LocalAI by @localai-bot in https://github.com/mudler/LocalAI/pull/10279
* fix(mlx): route vision-language models to the mlx-vlm backend by @localai-bot in https://github.com/mudler/LocalAI/pull/10274
* fix(darwin): fix vibevoice-cpp build linkage + fail-safe go backend packaging by @localai-bot in https://github.com/mudler/LocalAI/pull/10276
* fix(agents): emit chat event timestamps in milliseconds (#9867) by @aniruddh909 in https://github.com/mudler/LocalAI/pull/10243
* fix(realtime): keep transcription model on a language-only session.update by @localai-bot in https://github.com/mudler/LocalAI/pull/10295
* chore: :arrow_up: Update mudler/locate-anything.cpp to `92c1682da792c1e8a5dec91acc2be4b02c742ded` by @localai-bot in https://github.com/mudler/LocalAI/pull/10282
* fix(config): backend-gate the top_k=40 sampler default (#6632) by @localai-bot in https://github.com/mudler/LocalAI/pull/10285
* feat(gallery): add 60 piper TTS voices across 42 languages (Phase 2) by @localai-bot in https://github.com/mudler/LocalAI/pull/10296
* fix(deps): bump cogito to fix MCP image-result panic (#10101) by @localai-bot in https://github.com/mudler/LocalAI/pull/10294
* fix(neutts): pin torchaudio to match torch (fixes undefined symbol) (#9798) by @localai-bot in https://github.com/mudler/LocalAI/pull/10292
* fix(gallery): make opus a meta backend for platform auto-selection (#9813) by @localai-bot in https://github.com/mudler/LocalAI/pull/10291
* chore(model gallery): :robot: add 1 new models via gallery agent by @localai-bot in https://github.com/mudler/LocalAI/pull/10298
* fix(gallery): correct meta-backend definitions for platform auto-selection by @localai-bot in https://github.com/mudler/LocalAI/pull/10299
* chore(model gallery): :robot: add 1 new models via gallery agent by @localai-bot in https://github.com/mudler/LocalAI/pull/10302
* ci(darwin): build the ds4 backend for darwin/arm64 (metal) by @localai-bot in https://github.com/mudler/LocalAI/pull/10303
* fix(react-ui): stop Talk pipeline overflow and center collapsed-rail avatar by @localai-bot in https://github.com/mudler/LocalAI/pull/10305
* chore(model gallery): :robot: add 1 new models via gallery agent by @localai-bot in https://github.com/mudler/LocalAI/pull/10304
* fix(react-ui): make agent chat timestamps format-agnostic (#9867) by @localai-bot in https://github.com/mudler/LocalAI/pull/10290
* model: fix case-insensitive suffix matching and skip .bak files in ListFilesInModelPath by @pos-ei-don in https://github.com/mudler/LocalAI/pull/10306
* fix(xsysinfo): container-aware total RAM detection (cgroup/lxcfs) (#8059) by @localai-bot in https://github.com/mudler/LocalAI/pull/10288
* feat(distributed): declarative per-model scheduling via env/args by @localai-bot in https://github.com/mudler/LocalAI/pull/10308
* feat(sherpa-onnx): add Kokoro TTS + multilingual Piper voices by @localai-bot in https://github.com/mudler/LocalAI/pull/10309
* feat(omnivoice-cpp): add OmniVoice TTS backend (file + streaming, voice cloning + voice design) by @localai-bot in https://github.com/mudler/LocalAI/pull/10310
* feat(i18n): add Korean (ko) translation by @moduvoice in https://github.com/mudler/LocalAI/pull/10312
* feat(qwen3-tts-cpp): migrate to ServeurpersoCom/qwentts.cpp (streaming, speakers, voice design) by @localai-bot in https://github.com/mudler/LocalAI/pull/10316
* feat(realtime): gate realtime pipeline voice models behind voice recognition by @localai-bot in https://github.com/mudler/LocalAI/pull/10319
* chore: :arrow_up: Update vllm-project/vllm cu130 wheel to `0.23.0` by @localai-bot in https://github.com/mudler/LocalAI/pull/10314
* test(e2e): live-server voice-recognition gate test by @localai-bot in https://github.com/mudler/LocalAI/pull/10324

## New Contributors
* @aniruddh909 made their first contribution in https://github.com/mudler/LocalAI/pull/10243
* @moduvoice made their first contribution in https://github.com/mudler/LocalAI/pull/10312

**Full Changelog**: https://github.com/mudler/LocalAI/compare/v4.4.2...v4.4.3