whisper_edge 0.1.1

On-device speech-to-text for Flutter, powered by whisper.cpp. Offline transcription with timestamps, language detection and model downloads, on iOS and Android.

0.1.1

• WhisperModel: full ggerganov/whisper.cpp catalog (33 models) with exact published sizes — adds medium, large-v1/v2/v3, large-v3-turbo and all q5/q8 + .en quantized variants.
• WhisperModelDownloader.delete removes a downloaded model (and any stale .part file).
• Fixed englishOnly detection for quantized .en variants (e.g. ggml-base.en-q5_1.bin).

0.1.0

• Initial release: offline transcription on iOS and Android via whisper.cpp v1.9.1.
• WhisperTranscriber — isolate-based transcription with timestamps, language detection, translate-to-English, progress and cancellation.
• WhisperModelDownloader + WhisperModel catalog (official ggml models from Hugging Face).
• WAV decoding helpers (PCM 8/16/24/32-bit and float32, resampling, downmix).
• iOS: prebuilt Metal/Accelerate xcframework (iOS 16.4+). Android: built from source with the NDK (minSdk 24).