settings tour

Settings is a single scrolling column; every section also has a direct link (handy when a doc page points you somewhere). The quick tour:

• appearance — light, dark, or follow the system.
• language — the app's display language: English, Español or Nederlands, or automatic to follow your system. This is chrome only — transcription is always detected by ear, never tied to this, and your transcripts are never translated behind your back.
• microphone — default input device. As the card itself says: audio never leaves this device.
• model folders — extra folders to search for language models you already have (a model file or another Ollama library) so nothing downloads twice.
• keyboard — every shortcut, rebindable: click one, press the new keys, Esc cancels. Shortcuts can also be disabled outright.
• speakers / vocabulary / lenses — mirrors of the directory, glossary and library.
• entity types — which kinds of things (people, places, organisations…) the understand-step extracts from transcripts.
• compute — per-stage backend routing: On-device (default, private), yapyap Cloud, or bring-your-own OpenAI / Anthropic / OpenAI-compatible key.
• setup & license — re-run onboarding pieces; manage activation.