Speakr: features, limits & comparisons
Upload or record calls, separate speakers, keep searchable notes and inspect speaker stats; the server is now your problem.
murtaza-nasir.github.io ↗repo ↗
functionality and fit
Recorded license: AGPL-3.0.
- ASR via self-hosted WhisperX, OpenAI transcription, Mistral Voxtral, AssemblyAI, VibeVoice or custom endpoints; LLMs via any OpenAI-compatible endpoint including OpenRouter, OpenAI, Gemini, LocalAI or Ollama
- SQLite in the mounted instance directory, audio in mounted uploads or S3-compatible storage, and optional auto-exports such as Markdown and SRT files
- Download the example Docker Compose and .env, add transcription/text-model credentials or endpoints, set the admin account and run docker compose up -d
Costs and editions: A source license is separate from operating costs. Hosting, storage, external APIs and commercial support may require payment.
scope and migration checks
- It has no managed calendar bot or automatic conferencing integrations; capture, uptime and every external integration are yours to operate.
- It can centralize recordings and notes, but getting there requires a server and model endpoints, with less polished meeting automation, calendar integration and team administration than Supernormal.
- It is a self-hosted archive rather than a turnkey meeting bot, so capture and structured minutes depend on the ASR and LLM configuration.
recorded catalogue description · recorded source check 2026-08-11 · suitability checks are editorial assessments, not hands-on tests
licenseAGPL-3.0
runsyour server
installDownload the example Docker Compose and .env, add transcription/text-model credentials or endpoints, set the admin account and run docker compose up -d
enginesASR via self-hosted WhisperX, OpenAI transcription, Mistral Voxtral, AssemblyAI, VibeVoice or custom endpoints; LLMs via any OpenAI-compatible endpoint including OpenRouter, OpenAI, Gemini, LocalAI or Ollama
dataSQLite in the mounted instance directory, audio in mounted uploads or S3-compatible storage, and optional auto-exports such as Markdown and SRT files
setupDocker is mandatory for the easy path, and the operator must supply API keys or deploy a separate GPU ASR container plus reverse proxy/TLS
recorded source check 2026-08-11 · installation not tested
compare these app workflows
These are scoped comparisons, not guarantees that Speakr reproduces every paid feature. Follow each build guide for the implementation boundary and the collection for other options.
Upload or record calls, separate speakers, keep searchable notes and inspect speaker stats; the server is now your problem.
the verdict →A polished self-hosted call archive with transcription, summaries, search and chat; zero subscription, nonzero server.
the catchIt has no managed calendar bot or automatic conferencing integrations; capture, uptime and every external integration are yours to operate.
the verdict →all Fireflies.ai alternatives →Templates, summaries, tasks and searchable meetings in a self-hosted workspace.
the catchIt can centralize recordings and notes, but getting there requires a server and model endpoints, with less polished meeting automation, calendar integration and team administration than Supernormal.
the verdict →all Supernormal alternatives →Self-hosted transcription, templates, topics, tasks, search and chat in one polished web app.
the catchIt is a self-hosted archive rather than a turnkey meeting bot, so capture and structured minutes depend on the ASR and LLM configuration.
the verdict →all Sembly AI alternatives →Podcast repurposing with your own model endpoints instead of another monthly invoice.
the verdict →Transcribes meetings, summarizes them and lets you chat across the archive; the price is operating the stack yourself.
the verdict →Podcast repurposing with your own model endpoints instead of another monthly invoice.
the verdict →A self-hosted archive for recordings, transcripts, summaries and search; bring a server and preferably a GPU.
the catchIt can become a capable private archive, but requires server and model administration and is less turnkey than Notta's hosted upload, live transcription, collaboration and mobile access.
the verdict →all Notta alternatives →latest catalogue source check 2026-08-11 · recorded recurring plans: $1,597.8/yr (before operating costs) · no votes, no pay-to-list
sources and related categories
These are sources retained with the catalogue record. Their recorded check date is not a new verification of current terms.
- murtaza-nasir.github.io/speakr ↗
- github.com/murtaza-nasir/speakr ↗
- github.com/murtaza-nasir/speakr/blob/master/LICENSE ↗
- murtaza-nasir.github.io/speakr/getting-started ↗
- murtaza-nasir.github.io/speakr/getting-started/installation ↗
- murtaza-nasir.github.io/speakr/admin-guide/model-configuration ↗
- murtaza-nasir.github.io/speakr/admin-guide/storage ↗
- fireflies.ai ↗