v1.4.6

modelscope/FunASRv1.4.6Aug 29, 2026by github-actions[bot]

AI Summary

This Python package release bundles the llama.cpp runtime v0.2.3 and includes documentation updates, bug fixes for timestamp generation and real-time timeouts, and a new CLI feature for grouping subtitle cues.

Key Highlights

Fix for IndexError on empty segments in timestamp generation
Fix for false keepalive timeouts under load in realtime mode
Fix for Vulkan backend initialization failures
New CLI feature to group readable subtitle cues
Documentation updates for runtime versions and GPT-SoVITS

New Features

Subtitle cue grouping in the command-line interface
Improved stability for real-time inference under load
Guard against backend initialization failures

Full Release Notes

## What's Changed
* docs(site): publish FunASR 1.4.5 deployment release by @LauraGPT in https://github.com/modelscope/FunASR/pull/3535
* docs(site): publish GPT-SoVITS Fun-ASR runtime contract by @LauraGPT in https://github.com/modelscope/FunASR/pull/3537
* test(site): sync GPT-SoVITS public fixture by @LauraGPT in https://github.com/modelscope/FunASR/pull/3538
* fix: guard against IndexError on an empty segment in timestamp_sentence_en by @AmirF194 in https://github.com/modelscope/FunASR/pull/3540
* feat(cli): group readable subtitle cues by @LauraGPT in https://github.com/modelscope/FunASR/pull/3542
* fix(realtime): avoid false keepalive timeouts under load by @LauraGPT in https://github.com/modelscope/FunASR/pull/3533
* fix(vulkan): guard backend initialization failures by @LauraGPT in https://github.com/modelscope/FunASR/pull/3543
* Update product site for llama.cpp v0.2.2 by @LauraGPT in https://github.com/modelscope/FunASR/pull/3545
* fix(runtime): trace SenseVoice backend stages by @LauraGPT in https://github.com/modelscope/FunASR/pull/3546
* docs(runtime): publish v0.2.3 diagnostics by @LauraGPT in https://github.com/modelscope/FunASR/pull/3547
* docs(site): publish llama.cpp runtime v0.2.3 matrix by @LauraGPT in https://github.com/modelscope/FunASR/pull/3548
* chore(release): prepare 1.4.6 by @LauraGPT in https://github.com/modelscope/FunASR/pull/3549

## New Contributors
* @AmirF194 made their first contribution in https://github.com/modelscope/FunASR/pull/3540

**Full Changelog**: https://github.com/modelscope/FunASR/compare/v1.4.5...v1.4.6

<!-- funasr-runtime-downloads:start -->
## Runtime downloads

This Python release pairs with the current prebuilt llama.cpp / GGUF runtime release: [runtime-llamacpp-v0.2.3](https://github.com/modelscope/FunASR/releases/tag/runtime-llamacpp-v0.2.3).

The same verified runtime assets are attached directly to this Python release so users can find the package and self-contained `llama-funasr-*` binaries in one place.

| Platform | Asset | SHA-256 |
|---|---|---|
| Linux arm64 | [funasr-llamacpp-linux-arm64.tar.gz](https://github.com/modelscope/FunASR/releases/download/v1.4.6/funasr-llamacpp-linux-arm64.tar.gz) | `54abd8dbdcfc200a64a62b657f40d6e4c123c423988707058ce2771096f2921a` |
| Linux x64 AVX2 | [funasr-llamacpp-linux-x64-avx2.tar.gz](https://github.com/modelscope/FunASR/releases/download/v1.4.6/funasr-llamacpp-linux-x64-avx2.tar.gz) | `139616c8adf6b5b306cefcfb1813fabc0f4b7727d655d79102163c5c146df200` |
| Linux x64 Vulkan | [funasr-llamacpp-linux-x64-vulkan.tar.gz](https://github.com/modelscope/FunASR/releases/download/v1.4.6/funasr-llamacpp-linux-x64-vulkan.tar.gz) | `ea8eeb9e334598a59c4cdb2dba282855b9a3b909cd05684d30250914f418bf43` |
| Linux x64 portable | [funasr-llamacpp-linux-x64.tar.gz](https://github.com/modelscope/FunASR/releases/download/v1.4.6/funasr-llamacpp-linux-x64.tar.gz) | `5b7cf0c2339ab76d8a56594455b1607e02f1eab34392b2543b95a5273413f088` |
| macOS arm64 | [funasr-llamacpp-macos-arm64.tar.gz](https://github.com/modelscope/FunASR/releases/download/v1.4.6/funasr-llamacpp-macos-arm64.tar.gz) | `532da7810e9311b275e3dd2e6f693811421ab49881e0194a788a6db189df4000` |
| Windows x64 AVX2 | [funasr-llamacpp-windows-x64-avx2.zip](https://github.com/modelscope/FunASR/releases/download/v1.4.6/funasr-llamacpp-windows-x64-avx2.zip) | `55fa78e2a7522e54ead84532452afe40041a60014ba0261add611124897a7fe8` |
| Windows x64 CUDA | [funasr-llamacpp-windows-x64-cuda.zip](https://github.com/modelscope/FunASR/releases/download/v1.4.6/funasr-llamacpp-windows-x64-cuda.zip) | `ee25ed9cb4dff763f94de7db5812dd69fd7e748d2940dfbf82f29c408a4d419e` |
| Windows x64 Vulkan | [funasr-llamacpp-windows-x64-vulkan.zip](https://github.com/modelscope/FunASR/releases/download/v1.4.6/funasr-llamacpp-windows-x64-vulkan.zip) | `1c90daf2292fcf9041c14a8017be65a05acc8e42443ee0bbc8a8dfddb78cbdc7` |
| Windows x64 portable | [funasr-llamacpp-windows-x64.zip](https://github.com/modelscope/FunASR/releases/download/v1.4.6/funasr-llamacpp-windows-x64.zip) | `54520b705ffedfc2b3505f337bfff64d7890e28b660e78e7255b27e1dc34dc0e` |

Quick start: download one asset, unpack it, then run the bundled `download-funasr-model.sh <sensevoice|paraformer|nano>` helper and one of `llama-funasr-cli`, `llama-funasr-sensevoice`, or `llama-funasr-paraformer`.

For Python users, install from PyPI:

```bash
python -m pip install -U "funasr==1.4.6"
```
<!-- funasr-runtime-downloads:end -->