Local AI Cat personal markLocal AI Cat business markLocal AI Catkeep your paws off my datafor very serious cat business
← All modules

Local AI Cat module

Transcription (Whisper / SpeechAnalyzer)

Current shipping, testing, and product context for this part of Local AI Cat.

Status

Where it stands

Lane: StableStatus: LiveTesting: prov. +behavioral
Platforms
iOS · macOS
Release channels
Alpha · Beta · Main
Distribution
Indoor · Outdoor
  • AlphaiOS · macOS
  • BetaiOS · macOS
  • Outdoor · Direct downloadiOS · macOS
  • Indoor · App StoreiOS · macOS

Description

What it does

Core request/result models, request source/mode/translation flag, scheduler, scheduling priority, runtime lane resolver/OpenAI-model exclusion policy, Apple engine selection/fallback policy, concurrency profile/configuration/status, device capability detection helper, job service, Whisper language picker/display/request-parameter presentation, Settings speech-language display/code groups, Settings transcription model-card presentation, Transcribe playback/recording duration formatting, and the Whisper model catalog (sizes, WhisperKit variant/tokenizer manifests, device-availability policy, stored-selection normalization, preference-driven selection ordering) are packaged in LocalAITranscription; the app runtime imports the packaged types directly while supplying app-owned feature-flag/availability inputs. The pre-SpeechAnalyzer SFSpeechRecognizer lane is RETIRED (pkg 76): nothing selects it, its adapters are deleted, and appleSpeechLocal/appleSpeechRemote survive only as public engine= names that answer 400 naming a replacement. SpeechAnalyzer, Whisper, and FluidAudio cover the matrix. TERMINAL: only the WhisperKit/Apple SpeechAnalyzer live adapters, recording UI, model-selection UI layout, localized model presentation (display names/descriptions), the filesystem-bound WhisperModelAssets, history, and global-transcription shell remain — irreducible app-shell.

Headless Local API

Drive it over the loopback API

API parity: FullLogging: L0 · prov.

Logging / observability: provisional grade L0 — structured=0 print=42 sentry=0.

discover
present
execute
present
observe
present
configure
present
cancel/stop
present
reset
not applicable
transcriptionmodelDownloads
  • /v1/audio/transcriptions
  • /v1/local/model-categories
  • /v1/local/model-discovery
  • /v1/local/model-discovery/add
  • /v1/local/model-discovery/{model_id}
  • /v1/local/models
  • /v1/local/models/downloads
  • /v1/local/models/downloads/{download_id}
  • /v1/local/models/external
  • /v1/local/models/external/import

Media

See it in action

Walkthrough coming soonUI-test video · GIF fallback
A short walkthrough will appear here when the module's UI test media is ready.

Backlog / improvements

What comes next

No improvements are listed for this module yet.

History

How it got here

No change history has been published for this module yet.