v1.3.14

modelscope/FunASRv1.3.14Jul 13, 2026by github-actions[bot]

AI Summary

This release focuses on documentation updates and bug fixes, highlighting the Fun-ASR-Nano model as the flagship option and improving benchmark visibility.

Key Highlights

Updated README to feature Fun-ASR-Nano as the flagship model
Added benchmark data showing Fun-ASR-Nano at 340x speed
Fixed SyntaxError in whisper_lib/triton_ops.py
Fixed CIF crashes in Paraformer related to segments with no fired boundaries

Full Release Notes

## What's Changed
* docs(README): lead Quick Start with flagship Fun-ASR-Nano by @LauraGPT in https://github.com/modelscope/FunASR/pull/3019
* docs(model_selection): lead with flagship Fun-ASR-Nano by @LauraGPT in https://github.com/modelscope/FunASR/pull/3020
* docs(README): de-patchwork pass 1 — de-duplicate Quick Start by @LauraGPT in https://github.com/modelscope/FunASR/pull/3021
* docs(README): benchmark — show flagship Fun-ASR-Nano at its vLLM speed (340x) by @LauraGPT in https://github.com/modelscope/FunASR/pull/3023
* docs(README): add Chinese CER to the benchmark table by @LauraGPT in https://github.com/modelscope/FunASR/pull/3024
* docs(i18n README): align benchmark with canonical (add Chinese CER, flagship Nano 340x) by @LauraGPT in https://github.com/modelscope/FunASR/pull/3025
* fix(sense_voice): SyntaxError in whisper_lib/triton_ops.py by @LauraGPT in https://github.com/modelscope/FunASR/pull/3027
* fix(paraformer): CIF crashes (IndexError) on segments with no fired boundaries by @LauraGPT in https://github.com/modelscope/FunASR/pull/3028


**Full Changelog**: https://github.com/modelscope/FunASR/compare/v1.3.13...v1.3.14