Home/Speech & Audio/SenseVoice
SenseVoice logo

SenseVoice

FunAudioLLM/SenseVoice

Multilingual speech understanding toolkit with ASR, emotion recognition, and audio event detection.

GraphCanon updated 3w · GitHub synced 3w · 42 views this month

9.0k stars799 forksLast push 3w C Other

Decision brief

SenseVoice is ideal for applications needing fast multilingual ASR with emotion recognition and audio event detection in over 50 languages.

Good fit when

  • Projects requiring speed, with non-autoregressive models being up to 15 times faster than Whisper
  • Environments where multilingual support is crucial, catering more than 50 languages

Avoid when

  • Applications needing Python-based implementation since SenseVoice relies on C and PyTorch for its framework
  • Scenarios demanding the lowest level of inference latency as it may have higher overhead compared to simpler models

Observed Jul 17, 2026 · Source: enrich:decision_facts

Verify the decision

Maintenance and security

Full trust report
Maintenance
Very active (2d since push)
As of 3w
Provenance
Not a fork · Organization account
As of 3w
Security (OSV)
No criticals
As of 1mo

Public GitHub metadata and optional OSV scans. Signals, not a guarantee. Trust methodology.

Install

git clone https://github.com/FunAudioLLM/SenseVoice

Similar tools

Same-category neighbours. No typed graph edges are catalogued for this tool yet.

Evidence and technical details

Sourced facts, taxonomy, compatibility claims, README excerpt, and machine-readable endpoints.

Overview

SenseVoice is a comprehensive multilingual speech understanding software that supports automatic speech recognition (ASR), emotion recognition, and audio event detection in over 50 languages. It offers faster processing speed compared to Whisper while supporting non-autoregressive models.

Capability facts

Deploy
Self-host

Source: dockerfile:Dockerfile · Jul 30, 2026

Docker
Dockerfile present

Source: dockerfile:Dockerfile · Jul 30, 2026

Languages
c

Source: github.language · Jul 30, 2026

Categories

Compatibility

Sourced claims from the README excerpt - not unsourced marketing copy.

Python runtimePython

Source: README excerpt (regex_v1, Jul 30, 2026)

pip install -r requirements.txt
Source link

Tags

README

Requirements

pip install -r requirements.txt

SenseVoiceSmall examples and the composed FunASR diarization path require funasr>=1.3.26. If you installed this repository earlier, run pip install -U "funasr>=1.3.26" before retrying the demos.


pip3 install -U funasr funasr-onnx

from pathlib import Path from funasr_onnx import SenseVoiceSmall from funasr_onnx.utils.postprocess_utils import rich_transcription_postprocess

model_dir = "iic/SenseVoiceSmall"

model = SenseVoiceSmall(model_dir, batch_size=10, quantize=True)


Deployment with FastAPI

export SENSEVOICE_DEVICE=cuda:0
fastapi run --port 50000

Requirements

git clone https://github.com/modelscope/FunASR.git && cd FunASR
pip3 install -e ./

🐳 Docker Support

SenseVoice can be built and run using Docker to simplify setup, ensure reproducibility, and support both CPU and GPU inference.


Build with Docker

docker build -t sensevoice .

Docker Compose

Docker Compose provides an easier way to run SenseVoice with persistent model caching, networking etc.


License

  • Source code in this repository is licensed under the MIT License.
  • Model weights are distributed separately and follow the terms on each model card. The official SenseVoiceSmall model card links to the FunASR Model Open Source License Agreement; other artifacts and conversions may list different terms, so check their model cards before use.
  • The maintainers have provided an official SenseVoiceSmall license clarification for the FunASR Model Open Source License Agreement v1.1: Commercial use of the official SenseVoiceSmall weights is permitted when the model license is followed; Section 3 is a responsibility and risk disclaimer rather than an additional non-commercial restriction; and fine-tuned derivative weights may remain private. The Section 2.2 attribution and model-name requirements still apply. This clarification covers the official weights only, so check the terms for third-party conversions and bundled artifacts separately.

For agents

This page has a .md twin and JSON over the API.

Was this helpful?

Anonymous feedback helps us improve pages and translations.