FunASR

What is FunASR?

FunASR is the speech recognition toolkit from the speech lab of Alibaba DAMO Academy, published on ModelScope. It bundles voice activity detection, recognition (Paraformer, SenseVoice and other models), punctuation, timestamps and hotwords into ready-made services. The runtime SDK ships Docker images for an offline file transcription service (audio and video in, punctuated text out, hundreds of concurrent requests) and a real-time streaming service, both on CPU. A pip-installed funasr-server exposes an OpenAI-compatible transcription API.

Where it fits in a 10–200 person business

• Transcribe meeting recordings, customer calls or training videos in Chinese and English on your own server
• Add hotwords for product and brand names so transcripts get them right
• Give Dify, n8n or other tools a local OpenAI-compatible speech-to-text endpoint

These use cases are our editorial suggestions, not claims made by the project.

Can a business use FunASR? (license)

This is a plain-English summary of the license text, not legal advice. Third-party models you run with it have their own licenses.

Data & compliance: does FunASR phone home?

Checked against the project's own docs and repository on Nov 3, 2026. "Not documented" means we could not find the answer in official sources; it does not mean the feature is missing or that no data is sent. Third-party services you connect (model APIs, vector databases, search engines) have their own data policies. Compare all apps.

What hardware does FunASR need?

The offline file transcription tutorial lists x86 compute-type servers: 4 vCPU / 8 GB RAM handles about 32 concurrent requests, 16 vCPU / 32 GB about 64, and 64 vCPU / 128 GB about 200. No disk figure is published.

Source: github.com/modelscope/FunASR/blob/main/runtime/docs/SDK_tutorial.md

How much does it cost to host FunASR per month?

How to deploy FunASR

Method: Docker (offline file transcription runtime, CPU)

sudo docker pull registry.cn-hangzhou.aliyuncs.com/funasr_repo/funasr:funasr-runtime-sdk-cpu-0.4.7
sudo docker run -p 10095:10095 -it --privileged=true -v /root:/workspace/models registry.cn-hangzhou.aliyuncs.com/funasr_repo/funasr:funasr-runtime-sdk-cpu-0.4.7
# inside the container: cd /workspace/FunASR/runtime && nohup bash run_server.sh --download-model-dir /workspace/models --vad-dir damo/speech_fsmn_vad_zh-cn-16k-common-onnx --model-dir damo/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-onnx --punc-dir damo/punc_ct-transformer_cn-en-common-vocab471067-large-onnx --lm-dir damo/speech_ngram_lm_zh-cn-ai-wesp-fst --itn-dir thuduj12/fst_itn_zh --hotword /workspace/models/hotwords.txt > log.txt 2>&1 &

From the official instructions: github.com/modelscope/FunASR/blob/main/runtime/docs/SDK_advanced_guide_offline.md. Always check the source for the current version before you run it.

Known gotchas when self-hosting FunASR

• The OpenAI-compatible server has no authentication and no upload-size limit, and listens on 0.0.0.0 by default. Bind it to 127.0.0.1 and put TLS, auth and request limits in a gateway before sharing it. source
• The documented docker run uses --privileged=true and mounts a host folder (/root in the example) as /workspace/models; models are downloaded from ModelScope on first start. source
• Model weights are licensed separately from the MIT toolkit; third-party models in the model zoo (for example MOSS-Transcribe-Diarize) carry their own terms. source

Alternatives in Speech & translation

• whisper.cpp Fast C/C++ port of OpenAI Whisper speech recognition with a small HTTP transcription server.
• PDFMathTranslate PDF translator that keeps the original layout, formulas, charts and tables.
• Fish Speech Multilingual text-to-speech and voice cloning server (Fish Audio Drama 3 and S2-Pro) with an HTTP API and web UI.
• Meetily Desktop meeting recorder that transcribes and summarizes meetings locally.
• CosyVoice Multilingual LLM-based text-to-speech with zero-shot voice cloning, Chinese dialects and streaming output.
• LibreTranslate Self-hosted machine translation API with no dependency on Google or DeepL.
• Anarlog Local-first AI meeting notepad that records device audio without a bot joining the call.
• Speaches OpenAI-compatible speech-to-text and text-to-speech server (faster-whisper, Kokoro, Piper).
• Whisper ASR Webservice Whisper speech-to-text as a REST API in one Docker container.
• Scriberr Self-hosted web app that transcribes audio offline, with speaker labels and AI summaries.

Sources

• GitHub repository: github.com/modelscope/FunASR
• Official documentation, Hardware requirements: github.com/modelscope/FunASR/blob/main/runtime/docs/SDK_tutorial.md
• License text: github.com/modelscope/FunASR/blob/main/MODEL_LICENSE
• Deploy instructions, Gotcha 2: github.com/modelscope/FunASR/blob/main/runtime/docs/SDK_advanced_guide_offline.md
• Gotcha 1: github.com/modelscope/FunASR/blob/main/examples/openai_api/SECURITY.md
• Gotcha 3: github.com/modelscope/FunASR#model-zoo
• Hetzner Cloud pricing (cost estimate): www.hetzner.com/cloud/regular-performance
• GitHub stars, commits and releases: GitHub REST API, fetched Nov 3, 2026