NewYour coding agent can read the release notes before it upgrades.Set up the MCP server →
pub.dev · #1524 most downloaded on pub.dev
Speech recognition, speech synthesis, and speaker recognition using next-gen Kaldi with onnxruntime without Internet connection.
Last release 20 days ago
11 Sep 2026
Ships fairly regularly
a new release about every 2 weeks
Nearly every release is documented
notes for 60 of the last 60 stable releases
Nothing withdrawn
no release was ever pulled
2 years old
101 releases · first in 2024
Fix: export-onnx.py(expected all tensors to be on the same device)
Fix initializing TTS in Python.
One column per month.
Add new tts models for Latvia and Persian+English
Update AAR version in Android Java demo
Add missing changes about speaker identification demo for HarmonyOS
Fix building node-addon package
Add non-streaming ASR support for HarmonyOS.
sherpa_onnx.har for HarmonyOS (#1572)Support cross-compiling for HarmonyOS
Publish pre-built wheels for Python 3.13
Fix building node-addon for Windows x86.
Add Go API for offline punctuation models
Fix swift example for generating subtitles.
Add non-streaming ONNX models for Russian ASR
Add links to projects using sherpa-onnx.
Allow tokens and hotwords to be loaded from buffered string directly
Add VAD and keyword spotting for the Node package with WebAssembly
flutter: add lang, emotion, event to OfflineRecognizerResult
Add Pascal API for reading wave files
Fix splitting sentences for MeloTTS
Add Dart API for adding punctuations to text
Prefix all C API functions with SherpaOnnx
Fix the case when recognition results contain the symbol ". It caused issues when converting results to a json string.
". It caused
issues when converting results to a json string.Add Dart API for keyword spotter.
Support zh-en TTS model from MeloTTS.
Downgrade onnxruntime from v1.18.1 to v1.17.1
Update onnxruntime from v1.18.0 to v1.18.1
Update onnxruntime from 1.17.1 to 1.18.0
Add Flush to VAD so that the last speech segment can be detected. See also https://github.com/k2-fsa/sherpa-onnx/discussions/1077#discussioncomment-99
Support the iOS platform for Flutter.
Build sherpa-onnx into a single shared library.
Fix released packages. piper-phonemize was not included in v1.10.8.
Fix released packages. There should be a lib directory.
Nothing published for this version
* Fix passing C# string to C++
* Enable to stop TTS generation
* Add inverse text normalization
* Add TTS
* Publish with CI
* Fix path separator on Windows.
* Support specifying lib path.
* Initial release.
Your coding agent can read these notes before it upgrades. Set up the MCP server →