Whisper Tiny: Core ML encoder and full models

This repository contains the optional Apple Silicon encoder for Whisper Tiny in Glimpse. It is a companion to a Whisper GGUF or whisper.cpp ggml model, not a replacement for it: the model file still provides the decoder, tokenizer and metadata, and this package moves the audio encoder onto Core ML. Glimpse uses it for every Whisper Tiny quantization.

File	Download size	Purpose
`whisper-tiny-encoder.mlmodelc.zip`	15.0 MB	Compiled Core ML encoder

On Windows, Intel Macs, or anywhere without Core ML, use the model file on its own. The Core ML package requires Apple Silicon.

Full models

Complete model files for Windows, Intel Macs and any CPU/GPU path. They need no encoder.

File	Download size	Format
`whisper-tiny-Q8_0.gguf`	46.0 MB	Full GGUF, Q8_0
`whisper-tiny-F16.gguf`	80.1 MB	Full GGUF, F16
`ggml-tiny-q5_1.bin`	32.2 MB	whisper.cpp ggml, Q5_1
`ggml-tiny-q8_0.bin`	43.5 MB	whisper.cpp ggml, Q8_0
`ggml-tiny.bin`	77.7 MB	whisper.cpp ggml, F16

Provenance

The model originates from OpenAI Whisper Tiny, licensed Apache 2.0. The encoder was converted from whisper.cpp's F16 ggml-tiny.bin with scripts/convert-whisper-gguf-to-coreml.py from our transcribe.cpp fork, then zipped with ditto -c -k --keepParent.

The full GGUFs are unmodified copies of handy-computer/whisper-tiny-gguf at commit 1bd61cda, converted by handy-computer for transcribe.cpp (see NOTICE). The .bin files are unmodified copies of ggerganov/whisper.cpp at commit 5359861c, under OpenAI's MIT license for Whisper (LICENSE-MIT).

Artifact	SHA-256
Encoder ZIP	`35041e7f9f9c3e016bf1ff23819109c9dbd60cb387f97526fd44ba26c916bf52`
`whisper-tiny-Q8_0.gguf`	`325b9c7997cd1eff81ef709d55766565e71be696130cc3a3d444713798706834`
`whisper-tiny-F16.gguf`	`5b44043278b47d3b6e56fb16c6bc5bb0aa16f2e69086f4d67175ed0a30d6a987`
`ggml-tiny-q5_1.bin`	`818710568da3ca15689e31a743197b520007872ff9576237bda97bd1b469c3d7`
`ggml-tiny-q8_0.bin`	`c2085835d3f50733e2ff6e4b41ae8a2b8d8110461e18821b09a15c40c42d1cca`
`ggml-tiny.bin`	`be07e048e1e599ad46341c8d2a135645097a538221678b7acdd1b1919c6e1b21`

Runtime requirements

Extract the ZIP next to the model file, keeping the whisper-tiny-encoder.mlmodelc directory name. Glimpse-Speech looks for the companion beside the model and falls back to the model's own encoder when it isn't there.

The encoder follows Apple's Neural Engine transformer layout and computes in FP16. It takes one 30-second Whisper window. Core ML may run some operations on the CPU; this is not a guarantee of exclusive Neural Engine execution.

Glimpse

This encoder speeds up on-device dictation on Apple Silicon in Glimpse, a free, open-source dictation app for Mac and Windows. The source is on GitHub.

License