Starstruk Logo MarkStarstruk

v1.12.15

k2-fsa/sherpa-onnxv1.12.15Oct 22, 2025by csukuangfj

AI Summary

Version 1.12.15 introduces audio tagging APIs, expands TTS models (Piper, Parakeet TDT), fixes build issues for various platforms, and adds RKNN and Ascend NPU support for Paraformer models.

Key Highlights

  • Added CXX and C# API for audio tagging.
  • Added Parakeet TDT model for subtitles.
  • Added support for RK NPU (KWS and Paraformer).
  • Added support for Ascend NPU (Paraformer 910B).
  • Fixed Flutter and WASM build issues.

New Features

  • Audio tagging APIs
  • Parakeet TDT
  • RKNN support (KWS/Paraformer)
  • Ascend NPU support (Paraformer)
  • Piper Arabic TTS
  • WenetSpeech-Chuan support

Full Release Notes

## What's Changed
* exposing online punctuation model support in node-addon-api by @colourmebrad in https://github.com/k2-fsa/sherpa-onnx/pull/2609
* Fix building wheels by @csukuangfj in https://github.com/k2-fsa/sherpa-onnx/pull/2619
* Export one more Piper Arabic TTS model by @csukuangfj in https://github.com/k2-fsa/sherpa-onnx/pull/2623
* fix: hot update language for sencevoice by @o-u-p in https://github.com/k2-fsa/sherpa-onnx/pull/2627
* develop zipvoice c-api by @BarryKCL in https://github.com/k2-fsa/sherpa-onnx/pull/2628
* Add CI tests for Zipvoice Go API by @csukuangfj in https://github.com/k2-fsa/sherpa-onnx/pull/2630
* Remove hardcoded dithering value in NeMo transducer recognizer by @vsd-vector in https://github.com/k2-fsa/sherpa-onnx/pull/2639
* Reduce verbose output about reading lexicon for TTS by @csukuangfj in https://github.com/k2-fsa/sherpa-onnx/pull/2648
* Add Parakeet TDT model for generating subtitles by @csukuangfj in https://github.com/k2-fsa/sherpa-onnx/pull/2649
* Add more Piper TTS models by @csukuangfj in https://github.com/k2-fsa/sherpa-onnx/pull/2651
* Add CXX API for audio tagging by @csukuangfj in https://github.com/k2-fsa/sherpa-onnx/pull/2652
* Add C# API for audio tagging by @csukuangfj in https://github.com/k2-fsa/sherpa-onnx/pull/2653
* Support KWS + RKNN. by @csukuangfj in https://github.com/k2-fsa/sherpa-onnx/pull/2190
* Support https://github.com/ASLP-lab/WenetSpeech-Chuan by @csukuangfj in https://github.com/k2-fsa/sherpa-onnx/pull/2656
* Fix building for android by @csukuangfj in https://github.com/k2-fsa/sherpa-onnx/pull/2657
* Fix ios build script by @vovasty in https://github.com/k2-fsa/sherpa-onnx/pull/2645
* Update kaldi-native-fbank by @csukuangfj in https://github.com/k2-fsa/sherpa-onnx/pull/2659
* Add missing python class definitions for builds without TTS support by @ius in https://github.com/k2-fsa/sherpa-onnx/pull/2660
* Refactor JNI by @csukuangfj in https://github.com/k2-fsa/sherpa-onnx/pull/2661
* Remove jieba from kokoro and matcha tts. by @csukuangfj in https://github.com/k2-fsa/sherpa-onnx/pull/2662
* add flet_sherpa_onnx in readme by @SamYuan1990 in https://github.com/k2-fsa/sherpa-onnx/pull/2663
* Remove cppjieba by @csukuangfj in https://github.com/k2-fsa/sherpa-onnx/pull/2664
* Add phrase matcher to merge words into phrases for TTS. by @csukuangfj in https://github.com/k2-fsa/sherpa-onnx/pull/2668
* Limit number of tokens per sentence in MatchaTTS. by @csukuangfj in https://github.com/k2-fsa/sherpa-onnx/pull/2671
* Update README to include a ROS2 project using sherpa-onnx by @csukuangfj in https://github.com/k2-fsa/sherpa-onnx/pull/2672
* Fix building Flutter APPs by @csukuangfj in https://github.com/k2-fsa/sherpa-onnx/pull/2673
* Export Paraformer to RKNN by @csukuangfj in https://github.com/k2-fsa/sherpa-onnx/pull/2689
* Update README.md add achatbot-go Projects using sherpa-onnx link by @weedge in https://github.com/k2-fsa/sherpa-onnx/pull/2691
* Add CI to export Paraformer to RKNN by @csukuangfj in https://github.com/k2-fsa/sherpa-onnx/pull/2692
* Support MatchTTS with English and Chinese by @csukuangfj in https://github.com/k2-fsa/sherpa-onnx/pull/2695
* Export Paraformer ASR models from FunASR to Ascend NPU 910B by @csukuangfj in https://github.com/k2-fsa/sherpa-onnx/pull/2697
* Update README to include Ascend NPU by @csukuangfj in https://github.com/k2-fsa/sherpa-onnx/pull/2698
* Fix WASM (JS) after adding zipvoice. by @csukuangfj in https://github.com/k2-fsa/sherpa-onnx/pull/2702
* Export SenseVoice ASR models to Ascend NPU 910B by @csukuangfj in https://github.com/k2-fsa/sherpa-onnx/pull/2707
* Fix building for various language bindings after adding zipvoice by @csukuangfj in https://github.com/k2-fsa/sherpa-onnx/pull/2709
* Release v1.12.15 by @csukuangfj in https://github.com/k2-fsa/sherpa-onnx/pull/2710

## New Contributors
* @colourmebrad made their first contribution in https://github.com/k2-fsa/sherpa-onnx/pull/2609
* @o-u-p made their first contribution in https://github.com/k2-fsa/sherpa-onnx/pull/2627
* @BarryKCL made their first contribution in https://github.com/k2-fsa/sherpa-onnx/pull/2628
* @vovasty made their first contribution in https://github.com/k2-fsa/sherpa-onnx/pull/2645
* @ius made their first contribution in https://github.com/k2-fsa/sherpa-onnx/pull/2660
* @SamYuan1990 made their first contribution in https://github.com/k2-fsa/sherpa-onnx/pull/2663
* @weedge made their first contribution in https://github.com/k2-fsa/sherpa-onnx/pull/2691

**Full Changelog**: https://github.com/k2-fsa/sherpa-onnx/compare/v1.12.14...v1.12.15