Orca Mobile Makes Phone Dictation Desktop-Native
Orca Mobile pairs with the Orca desktop app so developers can dictate replies to AI agents from iOS or Android, using local speech models or optional OpenAI transcription. The beta also supports monitoring worktrees, reviewing terminal output, and sending agent follow-ups from a phone ([official docs](https://www.onorca.dev/docs/mobile)).
Orca’s strongest idea is treating the phone as a lightweight control surface while keeping compute and session state on the desktop. The tradeoff is that mobile convenience depends entirely on a reachable paired host.
- –Local options such as Parakeet, SenseVoice, Zipformer, and Whisper Tiny give developers privacy and language-specific flexibility ([speech model settings](https://www.onorca.dev/docs/settings)).
- –OpenAI transcription provides a fallback when smaller local models struggle, though it requires supplying an API key.
- –Dictation is directly connected to agent workflows rather than being a standalone mobile keyboard.
- –The desktop dependency is both the product’s privacy advantage and its main limitation: closing Orca can drop the mobile session.
- –Orca’s open-source, cross-platform agent environment makes this especially relevant for developers managing multiple coding agents ([GitHub](https://github.com/stablyai/orca)).
DISCOVERED
1h ago
2026-10-02
PUBLISHED
1h ago
2026-10-02
RELEVANCE
AUTHOR
orca_build