# SpeakoFlow

> Documentation for SpeakoFlow, the free and open-source local voice layer for your desktop: dictation, Generate with Flow, AI cleanup, the assistant, screen vision, memory, profiles, settings, and shortcuts for Windows, macOS, and Linux.

## Docs

- [Behavior & memory](https://www.speakoflow.com/docs/assistant/behavior.md): Control how much conversation context the SpeakoFlow assistant sends each turn, whether long chats get summarized instead of dropped, and how long its replies run.
- [Assistant panel](https://www.speakoflow.com/docs/assistant/panel.md): The SpeakoFlow assistant panel is a floating always-on-top chat you open with a hotkey. Learn its shortcuts, how to ask by voice or text, attach images and screen regions, and resize the panel.
- [Models and providers](https://www.speakoflow.com/docs/assistant/providers.md): Choose what powers the SpeakoFlow assistant: a private offline model that needs no API key, your own Ollama or LM Studio server, or a cloud provider with your own key. Covers the curated local models and the full provider list.
- [Screen vision](https://www.speakoflow.com/docs/assistant/screen-vision.md): Ask the SpeakoFlow assistant about what is on your screen. Covers the three screen access modes, when the frame is captured, image quality per provider, and the macOS Screen Recording permission.
- [Voice output (TTS)](https://www.speakoflow.com/docs/assistant/voice-output.md): Have the SpeakoFlow assistant read its replies aloud, free and local with Kokoro or through OpenAI-compatible, OpenRouter, ElevenLabs, or Azure AI Speech voices. Covers voices, precision, and playback speed.
- [Web search](https://www.speakoflow.com/docs/assistant/web-search.md): Let the SpeakoFlow assistant look things up when a question needs current facts. Covers the six snippet search providers, OpenRouter's built-in search, search depth, and how the model decides to search.
- [Dictation basics](https://www.speakoflow.com/docs/dictation/basics.md): Press a hotkey and talk, and SpeakoFlow types your words into any app on your machine. This page covers the dictation loop, the default Dictate shortcut on each platform, and the choice between Hold and Tap recording.
- [Custom words](https://www.speakoflow.com/docs/dictation/custom-words.md): Teach SpeakoFlow to spell names, products, and phrases your way so dictation stops writing "Coco row" when you say Kokoro. Covers where to add custom words and how to tune the correction threshold.
- [Hands-free recording](https://www.speakoflow.com/docs/dictation/hands-free.md): Set SpeakoFlow's recording behavior to Tap so one press starts recording and the next stops it. Covers how to stop or cancel a hands-free recording and why the mid-recording tap-to-lock gesture is inactive.
- [Live vs. batch transcription](https://www.speakoflow.com/docs/dictation/live-transcription.md): Streaming models show your words in SpeakoFlow's overlay while you speak, but text is still inserted once after you stop. This page lists which transcription models stream and how the overlay style switches streaming on and off.
- [Output & pasting](https://www.speakoflow.com/docs/dictation/output.md): How SpeakoFlow inserts transcribed text into the active app, and every setting that changes it: the six paste methods and their platform support, clipboard handling, auto submit, trailing spaces, and the Linux typing tools.
- [The recording overlay](https://www.speakoflow.com/docs/dictation/overlay.md): The small on-screen indicator that shows SpeakoFlow is listening. This page covers the Auto, None, Minimal, and Live styles, the per-platform position defaults, and why the overlay is hidden by default on Linux.
- [Text replacements and spoken emoji](https://www.speakoflow.com/docs/dictation/text-replacements.md): Map a spoken phrase to text you use often, and say the name of an emoji to insert it. Covers rule options, magic tokens like [date] and [time], the emoji grammar, and where both passes sit in the dictation pipeline. Both are off by default.
- [SpeakoFlow documentation](https://www.speakoflow.com/docs/index.md): SpeakoFlow is free voice dictation and an AI assistant for Windows, macOS, and Linux. Speech-to-text runs on your own machine. Start here to install it and dictate your first sentence.
- [Performance and memory](https://www.speakoflow.com/docs/models/acceleration.md): How SpeakoFlow uses your GPU for transcription and how long it keeps models in memory. Covers Metal, Vulkan, the ONNX backend setting, idle model unloading, the local assistant's context size, and keeping the microphone warm.
- [Custom models](https://www.speakoflow.com/docs/models/custom-models.md): Search Hugging Face from inside SpeakoFlow and download any GGUF model to use as your local assistant, including vision-capable ones. Covers choosing a quantization and what the Include screen vision toggle does.
- [Languages and translation](https://www.speakoflow.com/docs/models/languages.md): Choose the transcription language in SpeakoFlow, use autodetect, and translate speech to English on your own device. Includes which models cover which languages and which ones can translate.
- [Transcription models](https://www.speakoflow.com/docs/models/transcription-models.md): The full catalog of speech-to-text models SpeakoFlow can download and run locally, with size, language coverage, streaming, and translation for each. Also covers older models, downloads, and deleting models.
- [History](https://www.speakoflow.com/docs/personalize/history.md): History keeps your recent SpeakoFlow dictations, Flow generations, and assistant chats on your device so you can replay, copy, star, re-transcribe, continue, or delete them. Covers the retention settings and their defaults.
- [Personal memory](https://www.speakoflow.com/docs/personalize/memory.md): SpeakoFlow's optional on-device memory lets the assistant remember how you work between chats. It is off by default, and every note and summary is visible, editable, exportable, and erasable in Settings.
- [Profiles](https://www.speakoflow.com/docs/personalize/profiles.md): Profiles are task-focused assistant personas you switch between from the SpeakoFlow panel header. Compare the four built-in profiles and their reply lengths, and learn how to create, restore, and share your own.
- [About & license](https://www.speakoflow.com/docs/reference/about.md): What SpeakoFlow's About section shows, where it keeps your data and logs, the current known limitations including Apple Silicon only builds and no Linux .rpm, the roadmap, and the MIT license and credits.
- [Privacy](https://www.speakoflow.com/docs/reference/privacy.md): What SpeakoFlow sends and what it keeps on your device: speech-to-text runs locally, there is no telemetry and no account, web search and personal memory are off by default, and API keys go to your OS credential store.
- [Troubleshooting](https://www.speakoflow.com/docs/reference/troubleshooting.md): Fixes for the most common SpeakoFlow problems: the macOS damaged-app message, blocked microphone or accessibility permissions, text that will not type, Linux hotkey permissions and the floating overlay, and clipped or wrong pastes.
- [Command-line flags](https://www.speakoflow.com/docs/settings/cli.md): Every command-line flag and environment variable SpeakoFlow accepts on Windows, macOS, and Linux, including the remote-control flags that start, stop, or cancel a recording on an already-running instance.
- [Debug & advanced](https://www.speakoflow.com/docs/settings/debug.md): How to reveal SpeakoFlow's hidden Debug section, plus every setting in it: log level and where the log file is written, the word correction threshold, paste delay, and the extra recording buffer.
- [General settings](https://www.speakoflow.com/docs/settings/general.md): Every setting in SpeakoFlow's General section with its default: recording shortcuts and microphone, theme and text size, feedback sounds, the recording overlay, startup and tray behavior, and the experimental toggles.
- [Getting started](https://www.speakoflow.com/docs/start/getting-started.md): A two-minute SpeakoFlow quickstart: grant microphone and accessibility permissions, pick a local speech-to-text model, optionally add a local assistant model, and dictate your first sentence.
- [How SpeakoFlow works](https://www.speakoflow.com/docs/start/how-it-works.md): The mental model behind SpeakoFlow: one hotkey layer over the app you are already in, a dictation pipeline that runs entirely on your machine, and an assistant you can run offline, on your own server, or in the cloud.
- [Install SpeakoFlow](https://www.speakoflow.com/docs/start/install.md): How to install SpeakoFlow on Windows, macOS, or Linux, including which Linux package to pick and the one-time Terminal command that clears the macOS "app is damaged" message on unsigned builds.
- [Keyboard shortcuts](https://www.speakoflow.com/docs/start/keyboard-shortcuts.md): Every SpeakoFlow keyboard shortcut and its default on Windows, macOS, and Linux, plus how Hold and Tap recording differ and why tap-to-lock currently does nothing.
- [AI cleanup](https://www.speakoflow.com/docs/writing/ai-cleanup.md): AI cleanup runs a language model over a dictation to strip filler and fix grammar and punctuation, on its own Dictate and clean up shortcut. Covers cleanup strength, fixing misheard words, where the model runs, and the timeout.
- [Generate with Flow](https://www.speakoflow.com/docs/writing/generate-with-flow.md): Begin a dictation with the words Hey Flow and SpeakoFlow writes the finished email, reply, or draft and pastes it where you are typing. Covers turning it on, renaming the phrase, screen access, and what happens when generation fails.
- [Writing styles](https://www.speakoflow.com/docs/writing/writing-styles.md): A writing style controls how cleaned dictation sounds in SpeakoFlow. Compare the six built-in styles, including Formal, Casual, and Concise, and learn how to save your own.