← Reddit

I built Sipario: a free native Android TV player built on mpv. Plays everything, and actually remembers your language/subtitle settings.

Reddit · Aiml3ss · July 6, 2026
Sipario is a free native Android TV player built on mpv that retains user-selected language and subtitle settings across viewing sessions, addressing persistent issues where other media apps fail to preserve preferences and crash during playback. The application supports multiple codecs including HEVC and Dolby Vision, contains no content catalog or account requirements, and allows users to direct it toward their own sources such as Stremio addons, Jellyfill, Plex, or local files. Development is ongoing with the creator actively seeking beta testers for Play Store submission.

Detailed Analysis

Sipario represents a small but telling example of how AI-assisted development is enabling individual developers to build polished, technically sophisticated software that would traditionally require a team. The app, a free native Android TV media player built on the mpv engine, emerged from a solo developer's frustration with two persistent annoyances in existing apps like Stremio: media players failing to remember language and subtitle preferences across episodes, and ExoPlayer-based apps crashing on certain codecs with generic "can't continue playing" errors. Rather than tolerate these papercuts, the developer built a replacement, using Claude—including what the post refers to as "Fable 5"—as a thinking partner for architecture decisions and a significant contributor to the actual code, while the human developer retained control over direction, code review, and hands-on testing on real hardware.

The technical scope of the project is notable for a single-person effort. Sipario supports demanding codecs including HEVC, Dolby Vision, TrueHD, and DTS, works across Android TV and macOS, and includes a companion syncing feature that lets users send addon URLs from a phone to the TV app rather than typing them out with a remote control—a small but meaningful usability improvement for the notoriously input-constrained TV interface. The app follows a "bring your own content" model, integrating with Stremio-style addons, Jellyfin, Plex, and live playlists without hosting any content itself, avoiding the legal and business complexities of being a content aggregator while still solving the playback-quality problem that plagues many of these front-ends.

This case matters beyond the niche of media player software because it illustrates a broader shift in what solo developers and small teams can realistically ship. Historically, building a cross-platform native player with robust codec support, persistent state management, and remote-device syncing would have demanded either a team with specialized video engineering experience or months of solitary, error-prone work. The developer's own framing—that AI "writes a lot of it" while the human sets direction and verifies functionality—reflects a maturing division of labor in AI-assisted coding, where the developer acts more as an architect and quality gatekeeper than as a line-by-line implementer. This mirrors patterns seen across other Claude-assisted projects surfacing in developer communities: ambitious, previously team-scale software being built and maintained by individuals, with AI compressing the time between identifying a problem and shipping a working solution.

More broadly, Sipario fits into the growing trend of AI coding tools lowering the barrier to entry for building infrastructure-adjacent software—media players, self-hosted tools, and open ecosystem clients—that serve privacy- and control-conscious users who want alternatives to ad-supported or subscription-locked platforms. The project's free, no-ads, no-account model, combined with its reliance on user-supplied sources, appeals to a segment of tech-savvy consumers already using tools like Jellyfin and Plex, and its emergence as a community-shared, actively maintained beta (complete with a public bug-reporting process and Play Store beta-testing recruitment) shows how AI-assisted development is also changing the pace and openness with which independent developers can iterate and engage with early adopters.

Read original article →