repowiserepowise

App

murtaza-nasir / speakr

Speakr is a personal, self-hosted web application designed for transcribing audio recordings

★ 4,176stars on GitHubPythonAGPL-3.0

Claude Code, Cursor, Codex or VS Code. No key. You approve one sign-in in your browser.

Setup instructions on GitHub

Releases on GitHub (latest v0.11.3, Nov 21, 2026)

Repowise maps public repos from their source. Not affiliated with murtaza-nasir. What is Repowise?

github.com/murtaza-nasir/speakrIndexed at 074c490

What is speakr?

speakr is a personal, self-hosted web application for transcribing audio recordings and turning them into organized, searchable notes. It runs on your own infrastructure, so sensitive conversations stay under your control. Transcription is handled by a configured engine, such as self-hosted WhisperX or a hosted service, and the server is a Python app built on Flask with a JavaScript front end. Its code health is 6.1 out of 10 (Fair), ranked #209 of 4,000+ indexed repos. The last commit to speakr was on Aug 7, 2026, and 15 people have contributed.

As of commit 074c490, indexed Aug 22, 2026. What it is comes from a model reading the README; the rest is measured from the source and git history, not from the maintainer.

For agentsspeakr.md

Repo facts: type, license, health, rank, contributors
speakr facts
TypeApp
LicenseAGPL-3.0
Code health6.1 / 10 (Fair)
Health rank#209 of 4,000+ indexed repos
Last commitAug 7, 2026
Contributors15
Bus factor1
Size415 files, 116,309 lines

Explore speakr

Ask about this code in your agent

This connects Repowise to speakr's code. It does not install speakr.

Useful if you want to change or fork speakr: your agent can ask how it is put together and get answers that cite its files.

Ask about speakr

Answers come from speakr's README and source, with the files they cite. Two free answers a day without an account.

Want this for your own codebase? Map your repo free

About this page

Get a page like this for your own codebase

Repowise built this map of murtaza-nasir/speakr from its source and git history. It parses every symbol, draws the dependency graph, scores each file's health from complexity, duplication, coverage and churn, finds hotspots and owners in the history, and writes documentation for every module. Your coding agent reads the same index over MCP, so its answers cite real files.

Free for 2 public repos, with a page like this one. No card. See how Repowise works

Prefer local? pip install repowise && repowise init indexes your own checkout, no account needed. Star Repowise on GitHub

murtaza-nasir/speakr FAQ

What is speakr?

speakr is a personal, self-hosted web application for transcribing audio recordings. It turns recordings into organized, searchable notes and runs on your own infrastructure.

How does speakr work?

speakr accepts audio from the browser or from uploaded files, sends it to a configured transcription engine, and builds summaries, events and chat on the resulting transcript. The Flask entry point is src/app.py.

What language is speakr written in?

speakr is written mainly in Python, with JavaScript for the front end and some shell scripts and Dockerfiles.

How is the speakr code organized?

speakr keeps server code in src, front-end code in static and templates, tests in tests, and documentation in docs. The entry point is src/app.py.

Who is speakr for?

speakr is built for privacy-conscious groups and individuals who want their sensitive conversations to stay on their own infrastructure.

What license does speakr use?

speakr is licensed under AGPL-3.0.

Code health

6.1out of 10Fair

Hotspots average 4.5, the place fixes pay off most.

On the leaderboards:#60 of 71 Python repos

Key features of speakr

  • Records from a microphone, the computer's system or browser-tab audio, or both mixed together, and accepts drag-and-drop audio files.
  • Transcribes recordings with self-hosted WhisperX, OpenAI, Mistral, AssemblyAI, or a custom ASR webservice.
  • Labels speakers with speaker diarization and recognizes the same person across recordings through voice profiles.
  • Generates summaries and extracts action items and calendar-worthy events, with prompts that can be customized per recording, tag, or folder.
  • Lets you chat with a single recording or search and ask questions across the whole library in Inquire Mode.
  • Organizes recordings with folders, smart tags, retention policies, and automated export to templated files.

How murtaza-nasir/speakr works

Audio enters through the web interface, which records from a microphone, system or browser-tab audio, or accepts uploaded files, and the capture logic lives in static/js/modules/composables/audio.js. The front end is driven by static/js/app.modular.js, which coordinates the interface modules. Requests reach the server through src/app.py, the entry point that starts the Flask application. The recording endpoints in src/api/recordings.py handle the API calls that create, read and update recordings. Templates in templates render the pages, and the tests directory holds test files such as test_webhooks.py that cover the server code. Transcription is delegated to the configured engine, and the resulting transcript is used for summaries, event extraction and chat.

The codebase has 415 files and 3,459 symbols in 7 modules, led by Python, JavaScript and Shell. Code health is 6.1 out of 10, rated fair.