AI tools, scored for your job
Get paid to train AI
🤝 You're in the Free & Open Source library — community-built tools you can use without paying. A different concept from our scored AI directory.
Home › Free & Open Source › Therapists › speechlib
✦ Free & open sourceTherapists

speechlib

Unifies speaker diarization, transcription, and recognition into a single pipeline.

Unified Speech Processing Pipeline

Combine diarization, transcription, and speaker recognition with actual names and time tags.

Free forever · open source · MIT
✓ Needs IT✓ Browser-based✓ Real-time collaboration
5.8
ZEKAI
Powerful Python Audio Pipeline
Free & open source · Therapists
Ease of setup4
Usefulness7
Maintenance7
Community5
★ 268
GITHUB
Apr 2026
UPDATED
MIT
LICENSE
🛠 Don't want to self-host? We can set it up →
⚡ Quick answer

Speechlib is a Python library that merges speaker diarization, ASR backends, and speaker recognition into a single pipeline for audio conversations. It suits developers and clinicians building local audio processing workflows who need accurate speaker labeling with time tags.

Runs on
Cross-platform
Hosting
Self-host
Languages
Python
License
MIT

What it is

The library integrates tools like PyAnnote, Faster-Whisper, and SpeechBrain to process audio files and generate transcripts containing actual speaker names and precise time tags. It also includes audio preprocessing functions and supports custom plug-and-play backends for ASR, diarization, and recognition.

Who it's for

It suits Python developers and mental health tech builders who need programmatic control over local audio transcription pipelines with known speaker mapping. It does not suit non-technical practitioners looking for a ready-to-use desktop application or a cloud service without writing code.

Key features

Unified Pipeline Processing

Combines diarization, ASR, and speaker recognition into a single script.

Multiple ASR Backends

Supports Faster-Whisper, OpenAI Whisper, HuggingFace, and AssemblyAI.

Known Speaker Recognition

Maps diarization tags to real names using reference audio subfolders.

Custom Backend Subclassing

Allows you to plug in custom diarizers, recognizers, or ASR models.

How to start

Up and running in 3 steps

1
Install the library via pip with CPU or GPU-supported torch dependencies
2
Set up your environment variables including a HuggingFace access token for diarization
3
Initialize the Pipeline with your chosen diarization and ASR backends to process audio files
Real talk

Our honest verdict

What's great

  • Unifies diarization, transcription, and speaker recognition into one workflow
  • Supports multiple ASR backends including Faster-Whisper and AssemblyAI
  • Maps anonymous speaker tags to real names using reference voice folders
  • Provides abstract base classes for plugging in custom backends

Where it falls short

  • Requires writing Python code and handling Python environment dependencies
  • Needs an external HuggingFace token and accepted terms for PyAnnote models
  • Requires ffmpeg installed on your system and proper GPU drivers for CUDA execution
  • Windows users must run their IDE as administrator to avoid specific privilege errors
Cost

Pricing & license

Free under its license, plus any API costs if you choose to use cloud-based transcription providers like AssemblyAI. The only other cost is running it yourself on local hardware or GPU compute.

Paid work

Open remote AI-training roles

880 open remote AI-training roles, updated November 16, 2026; most pay $20–$175 an hour. Zekai lists them free; the application goes to the employer’s platform.

See all open roles →
Compare

Alternatives

🎁 Free & open-source alternatives
Get more

Use it, or have us set it up

📄 Free PDF: speechlib for Therapists
A plain-language guide to getting your first result. Sent to your inbox.
🔔 64 people want a full video tutorial — add your vote
🛠 Don't want to set it up yourself?
We'll install, host & configure speechlib for you — and train your team.
Get it done for you ↗
Questions

FAQ

Can I use local custom fine-tuned Whisper models?

Yes, you can use CustomWhisperASR to load local fine-tuned model checkpoints.

How does speechlib handle unknown speakers?

If you do not provide a voices folder, speakers are automatically labeled as anonymous tags like SPEAKER_00 and SPEAKER_01.

Radar

Latest AI news

✉️ The weekly free toolkit
The best new free & open-source tools for Therapists — one short email a week, in your language.
Free · 1 email/week · unsubscribe anytime
Get paid to train AI →

See Zekai first in Google