An hour of audio transcribed in about 13s on MacBook Air

# Embedding Gemma Transcribe Source: [https://eg2.haku.sh/](https://eg2.haku.sh/) Google's EmbeddingGemma 2 is an embedding model\. We turned it into a speech\-to\-text model:*an hour of audio in about 13 seconds on a base MacBook Air*, entirely on device\. [Download for Mac \(free\)](https://eg2.haku.sh/download)v1\.6 · macOS 14\+ · Apple Silicon models download on first launch \(715 MB\) ## Can an embedding model transcribe speech? EmbeddingGemma 2 is a multimodal embedding model: text, images, audio and video in one shared space\. Say a word out loud and its audio lands close to the written word\. So we tried reading words straight out of that space\. ## How it works 1. EmbeddingGemma 2's audio encoder stays exactly as it is\. 2. A small reader moves each short slice of audio into the model's text space\. 3. The closest of ~229,000 English words wins, and the slices become a transcript\. 4. A tiny extra model adds punctuation and capitals\. Only the reader is trained: under two hours on 8×H100, on openly licensed English speech\. The base model was never trained for speech\-to\-text\. ## How fast? ModelRuns onMacSpeedEmbedding Gemma TranscribeNeural Enginebase M3 MacBook Air, 8 GB288×1 hour in 12\.6 sIBM Granite Speech 5\.0 TurboCTCGPU \(MLX\)base M3 MacBook Air, 8 GB~150×1 hour in 25\.7 sNVIDIA Parakeet TDT 0\.6B v3 \(FluidAudio\)Neural EngineM4 Pro156×NVIDIA Parakeet TDT 0\.6B v3 \(FluidAudio\)Neural EngineM5 Pro335×Phonon\-2, a Parakeet v3 fine\-tune \(FluidAudio\)Neural EngineM5 Pro480× × real time: seconds of audio per second of processing\. Rows 1–2: our tests on the same base M3 MacBook Air\. Rows 3–5:[FluidAudio's published benchmarks](https://github.com/FluidInference/FluidAudio/blob/main/Documentation/Benchmarks.md)on larger, newer Macs and different audio\. A rough guide, not a like\-for\-like ranking\. ## Where it stands Its word error rate is about 5% right now, against 2–3% for leading speech models\. We believe a better\-trained reader can close that gap; that's the next iteration\. ## Use it Drop in an audio file, or hold Right Option, speak and release to dictate into any app\. It runs on the Mac's Neural Engine; your audio never leaves the machine\.