--- title: "ChangelogChangelog" description: "Release notes for Detta, the fermion-research runtime, Phonon-2 Core ML and the Fermion Research models in one feed, newest first." canonical: "https://www.fermionresearch.com/changelog/" source: "Fermion Research" --- # Changelog Every release across Detta, the Phonon runtime, Core ML and our models, newest first. ## October 2026 October 6, 2026 · Phonon runtime fermion-research 0.2.10 ### [Faster on Intel and AMD, with segment timestamps](#fermion-research-0-2-10) - Phonon-2 starts much faster and decodes faster on Intel and AMD processors, on Linux and on Intel Macs, and uses less memory while it runs. Transcripts are unchanged, byte for byte. - `segments` gives the start and end of every sentence or pause-sized stretch of speech, in `--json` and in the server's `verbose_json`, in the shape OpenAI clients already read. - `pip install fermion-research` is now the whole setup on Linux, Windows and Intel Macs. On Apple silicon, `pip install "fermion-research[mlx]"` adds the MLX runtime. - The CUDA image `phonon-cuda:1.0.7` returns segments and word timestamps, and the CPU image `phonon-cpu:2.0.8` runs this release on amd64 and arm64. [Release notes](https://github.com/fermionresearch/phonon/releases/tag/v0.2.10) · [PyPI](https://pypi.org/project/fermion-research/0.2.10/) · [Docs](https://www.fermionresearch.com/docs/changelog/) October 6, 2026 · Detta Detta 1.0.26 ### [Detta now runs on the Neural Engine](#detta-1-0-26) - Speech is recognised on your Mac's Neural Engine. Text appears faster after you speak, with a median of 0.26 seconds from key-up to text, down from 0.33, at the same accuracy. - Detta holds about a sixth of the memory it did while it waits for you to talk. - Your dictionary now steers recognition while it decodes, so names and terms you add are recognised more reliably. - Long recordings are polished from the first word to the last and keep every sentence, and each word keeps its start and end time as the recording plays. - The first time it runs, Detta lets you know while it sets up the Neural Engine for your Mac. Dictation works within about a minute. [Read the post](https://www.fermionresearch.com/research/detta-neural-engine/) · [Download](https://www.fermionresearch.com/products/detta/) · [Release notes](https://www.fermionresearch.com/Detta-1.0.26.html) October 6, 2026 · Core ML phonon-coreml 1.1.1 ### [Any audio file, with word timings](#phonon-coreml-1-1-1) - The command-line tool reads any sample rate and the common audio formats, m4a included. - `--words` prints every word on its own line with its start and end time. - A clearer notice on the first run, and clearer error messages. [Release](https://github.com/fermionresearch/phonon-coreml/releases/tag/1.1.1) · [Run it](https://www.fermionresearch.com/models/phonon-2/#coreml) October 4, 2026 · Phonon runtime fermion-research 0.2.9 ### [Hotwords, and Phonon-2 on Intel Macs](#fermion-research-0-2-9) - Hotwords on every Phonon-2 engine. Give it up to 25 names and terms and it favours them whenever the audio is close. Without hotwords, every transcript is unchanged. - `--hotwords` on the command line, `hotwords=` in Python, and a `hotwords` field on the server, which also reads the OpenAI `prompt` field as a vocabulary list. - Hotwords work on Apple silicon, on Linux, Windows and macOS CPUs, and in the CUDA image `phonon-cuda:1.0.6`. - Intel Macs run Phonon-2 on the CPU engine (0.2.8, first published in this release). [Release notes](https://github.com/fermionresearch/phonon/releases/tag/v0.2.9) · [Hotwords guide](https://github.com/fermionresearch/phonon/blob/main/docs/hotwords.md) · [Docs](https://www.fermionresearch.com/docs/changelog/) October 4, 2026 · Core ML Phonon-2 Core ML · phonon-coreml 1.1.0 ### [Phonon-2 on the Apple Neural Engine](#phonon-2-coreml) - Phonon-2 for the Neural Engine, at the accuracy of the reference engine, 5.21 % average word error on the Open ASR Leaderboard's seven English test sets. - On an M5 MacBook Air an hour of speech becomes text in 6 seconds, and the GPU stays free. - Windows from 5 to 35 seconds share one set of weights, and every word carries its start and end time. - A Swift package and command-line tool for macOS 15 and iOS 18 or later, and a Python runner. [Hugging Face](https://huggingface.co/FermionResearch/Phonon-2-CoreML) · [GitHub](https://github.com/fermionresearch/phonon-coreml) · [Run it](https://www.fermionresearch.com/models/phonon-2/#coreml) October 3, 2026 · Detta Detta 1.0.24 ### [Stays out of your way](#detta-1-0-24) - The app you are dictating into keeps focus while Detta works from the menu bar. - Check for Updates in the menu, and a notice you will see when a new version is ready. - Report a problem from inside Detta. What you wrote is kept if it cannot be sent. - Performance improvements and bug fixes. [Release notes](https://www.fermionresearch.com/Detta-1.0.24.html) · [Download](https://www.fermionresearch.com/products/detta/) October 2, 2026 · Detta Detta 1.0.23 ### [Subtitles from any recording](#detta-1-0-23) - Save a file transcription as subtitles, in SRT or WebVTT, or as text with timestamps. - Performance improvements and bug fixes. [Release notes](https://www.fermionresearch.com/Detta-1.0.23.html) · [Download](https://www.fermionresearch.com/products/detta/) October 1, 2026 · Phonon runtime fermion-research 0.2.7 ### [Word timestamps](#fermion-research-0-2-7) - `fermion transcribe phonon-2 clip.wav --json` returns every word with its start and end in seconds, placed correctly on long audio. - `fermion serve` honours `timestamp_granularities`, so `verbose_json` carries words and segments. - Both Phonon-2 engines return timings, on Apple silicon and on CPUs. Transcripts are unchanged. [PyPI](https://pypi.org/project/fermion-research/0.2.7/) · [Docs](https://www.fermionresearch.com/docs/changelog/) October 1, 2026 · Phonon runtime fermion-research 0.2.6 ### [Clearer messages](#fermion-research-0-2-6) - `fermion serve` with a name that is not a published model stops in one line, before any network request. - `fermion transcribe` checks the audio file exists before it loads a model. - `fermion models` and the help texts show each command with its model. [PyPI](https://pypi.org/project/fermion-research/0.2.6/) · [Docs](https://www.fermionresearch.com/docs/changelog/) October 1, 2026 · Phonon runtime fermion-research 0.2.5 ### [The right engine for every CPU](#fermion-research-0-2-5) - The CPU engine reads your processor's features and picks the matching tier, from AVX2, AVX-512 and AMX on x86-64 to dotprod, i8mm and SVE on 64-bit Arm. - Raspberry Pi 3 and 4, other Arm boards without dotprod, and x86-64 processors without AVX2 run Phonon-2 on a baseline tier. - `fermion describe` shows the features found, the tier chosen and the binaries loaded, with `--json` for scripts. [PyPI](https://pypi.org/project/fermion-research/0.2.5/) · [Docs](https://www.fermionresearch.com/docs/changelog/) ## September 2026 September 30, 2026 · Phonon runtime fermion-research 0.2.4 ### [Every command names its model](#fermion-research-0-2-4) - Commands name their model, as in `fermion transcribe phonon-2 meeting.wav`, and `phonon`, `phonon-2` and `phonon-1` each run a fixed model. - `fermion serve` takes `--unix-socket` for an owner-only local socket and `--threads` for the CPU engine. - The CPU and CUDA containers take the model first and keep it in a cache volume. [PyPI](https://pypi.org/project/fermion-research/0.2.4/) · [Docs](https://www.fermionresearch.com/docs/changelog/) September 30, 2026 · Detta Detta 1.0.21 ### [More complete long recordings](#detta-1-0-21) - Transcribe a File cuts long recordings at natural pauses, so more of every recording comes back as text. - Long dictations finish faster. - Detta opens at login on new installs. - If the speech engine cannot start, Detta says why and lets you copy the details for support. [Release notes](https://www.fermionresearch.com/Detta-1.0.21.html) · [Download](https://www.fermionresearch.com/products/detta/) September 30, 2026 · Detta Detta 1.0.14 to 1.0.20 ### [Your keys, your way](#detta-1-0-17) - Choose any key or key combination as your hold key (1.0.14). - Double-tap to lock can use its own key or key combination (1.0.16). - Set more than one hold key (1.0.17). [Release notes](https://www.fermionresearch.com/Detta-1.0.17.html) · [Download](https://www.fermionresearch.com/products/detta/) September 29, 2026 · Detta Detta 1.0 ### [Detta, dictation for the Mac](#detta-1-0) - Hold a key and talk. Detta types what you say where your cursor is, in any app, punctuated and capitalised. - Phonon-2 runs on the Mac itself, and Detta removes ums, repeated words and false starts. - Quick notes keep their recordings, and Transcribe a File turns wav, mp3, m4a, flac, mp4, mov and more into text. - Free for Apple silicon Macs, and it keeps itself up to date. [Read the post](https://www.fermionresearch.com/research/detta/) · [Download](https://www.fermionresearch.com/products/detta/) September 29, 2026 · Models Phonon-2 · fermion-research 0.2.0 ### [Introducing Phonon-2](#phonon-2) - An open English speech recognition model in a 164 MB download, averaging 5.21 % word error on the Open ASR Leaderboard's seven test sets. - 174 times realtime through MLX on an M5 MacBook Air, and 143 times on eight Zen 5 cores. - fermion-research 0.2.0 makes it the default speech model on Apple silicon and on CPUs across macOS, Linux and Windows, with CPU and CUDA container images. - Weights released under CC-BY-4.0. [Read the post](https://www.fermionresearch.com/research/phonon-2/) · [Model](https://www.fermionresearch.com/models/phonon-2/) · [Hugging Face](https://huggingface.co/FermionResearch/Phonon-2) September 2, 2026 · Phonon runtime fermion-research 0.1.23 ### [Recordings of any length](#fermion-research-0-1-23) - Long recordings are cut into windows at pauses and transcribed in full, from the command line and from the server. - `--json` adds the start and end of each segment, and `verbose_json` on the server fills `segments`. [PyPI](https://pypi.org/project/fermion-research/0.1.23/) · [Docs](https://www.fermionresearch.com/docs/changelog/) ## August 2026 August 28, 2026 · Models Phonon-1 · Phonon-1 Micro ### [Introducing Phonon-1](#phonon-1) - An open English speech recognition model in a 415 MB download that runs on a laptop or a datacenter GPU. - It transcribes an hour of audio in about two and a half minutes. - Phonon-1 Micro, the smallest model of the family, joins it. [Read the post](https://www.fermionresearch.com/research/phonon-1/) · [Phonon-1](https://www.fermionresearch.com/models/phonon-1/) · [Phonon-1 Micro](https://www.fermionresearch.com/models/phonon-1-micro/) ## July 2026 July 27, 2026 · Models Neutrino-1 8B · 0.6B · 0.6B-Chat ### [Introducing the Neutrino-1 models](#neutrino-1) - Three open-weight language models trained for a compact ternary format. - Neutrino-1 8B scores 72.1 on MMLU from a 2.56 GB download. - `pip install fermion-research` runs them on CUDA, Apple silicon and x86, with a local OpenAI-compatible server and a llama.cpp fork. - The 0.6B model also serves as the draft for speculative decoding. [Read the post](https://www.fermionresearch.com/research/neutrino-8b/) · [Models](https://www.fermionresearch.com/models/) Source: [https://www.fermionresearch.com/changelog/](https://www.fermionresearch.com/changelog/)