Show HN: Kyoo – Self-hosted media browser (Jellyfin/Plex alternative)

Original Article ↗ Hacker News Discussion ↗

Overall reception

  • UI described as very pretty, clean, and modern; demo feels fast and responsive.
  • Time skipping/seek performance draws particular praise vs Plex/Jellyfin/Emby.
  • Several users say they’re glad to see more options as Plex changes direction.

Client support & casting

  • Currently Android and web clients only; there is an APK.
  • No Chromecast/Google TV or AirPlay integration yet; Chromecast support is planned within ~6 months, Google TV within ~1 year.
  • iOS/tvOS are unlikely soon due to Apple’s $100/year dev fee and lack of Apple hardware for testing, though the React Native codebase should in theory help.
  • Some users see lack of TV and iOS apps as a blocker and suggest donations to fund development.

Library management philosophy

  • Multiple users complain Jellyfin is too opinionated about file/folder naming and requires renaming or manual tagging.
  • Kyoo’s stated goal is “zero renaming”: it should work directly on a download folder and handle messy names (including typical fansub-style anime names).
  • Questions remain about how well it builds hierarchical collections from deeply nested or mixed folder structures; details are unclear.

Playback, transcoding & performance

  • Kyoo uses ffmpeg on small segments and spins up new processes when seeking far away, aiming for instant seeks without waiting for a long-running transcoder.
  • Users who heavily seek/skips say this solves a major pain point vs existing servers.
  • Direct play is constrained by what the client (especially browsers) supports; audio-format support is essentially “whatever the client can handle.”

Architecture & technology choices

  • Backend uses C#/.NET; other components use Python, Go, and TypeScript.
  • Postgres, RabbitMQ, and Meilisearch are used; RabbitMQ coordinates services (work queues, websocket communication, replica sync).
  • Some consider multiple containers and dependencies overkill for self-hosters; others defend specialized services for search and distributed transcoding.
  • There’s mild debate about .NET’s tooling and ecosystem, especially on non-Windows platforms and BSD.

Scope limitations (music, books, DVR)

  • Project is intentionally focused on movies/series/anime.
  • No plans for ebooks or music; they’re deemed different enough to require substantial extra work.
  • No OTA/DVR support; the maintainer was unfamiliar with the concept and hasn’t planned for tuner-based recording.

Documentation & presentation

  • Several users feel the README reads like LLM-generated marketing copy: wordy and vague in places.
  • Suggestions to tighten language and better explain differentiators.
  • Some concern over using copyrighted movie imagery in screenshots; maintainer notes those are on a removable branch.