textream

Textream is a free, open-source teleprompter designed for Mac, iPhone, and iPad, aimed at streamers, interviewers, presenters, and podcasters. It offers three reading modes: word tracking (highlights each word as you speak), classic (constant-speed auto-scroll), and voice-activated (scrolls while you talk). The Mac app displays text in a Dynamic Island-style overlay, a draggable floating window, or fullscreen on a Sidecar iPad. The iOS app adds near-camera reading and presenter-only camera recording. Key features include: - **Mirror mode** for teleprompter glass, flipping the prompt horizontally, vertically, or both in landscape orientation, with a compact button for live mirror, position, text size, speed, and playback controls. - **Eye-contact reading position** — the active line sits just below the front camera by default, or can be centered. - **Optional camera while reading** — show or hide the camera behind the prompt. - **Selfie video recording** — captures camera video and microphone audio, with front/back camera switching, saved to Photos with a recordings gallery. - **Presenter-only prompt** — the scrolling prompt is visible on screen but not burned into the saved video. - **Matching fonts** — Sans, Serif, Mono, and OpenDyslexic on both platforms. - **Speech recognition** — on-device or Apple-processed depending on Mac, language, and system availability; no microphone needed for classic mode. - **Display options** — pinned to notch, floating window, or fullscreen on a selected display; adjustable text area height (100–400 px) and font sizes (XS to XL). - **Hide from screen share** — prevents the overlay from appearing in screen recordings or video calls. - **Remote connection** — view the teleprompter on any device via a local network browser (HTTP + WebSocket), with QR code access, real-time sync, and no app needed on the remote device. - **Director mode** — a director can write, edit, and push scripts in real time from any browser, with a dedicated HTTP + WebSocket server (default port 7575), live text editing, read-locked highlighting, and real-time state streaming at 10 Hz. - **PowerPoint notes import** — drop a .pptx file to extract presenter notes as pages. - **Multi-page support** — navigate between pages with automatic advance. - **Live waveform** — visual voice activity indicator. - **Privacy** — no accounts, ads, analytics, or Textream-operated cloud service; speech recognition may send audio to Apple for processing (see privacy policy). - **Open source** — MIT licensed. For developers, the Director Mode API exposes an HTTP server and WebSocket server on the local network. The built-in web UI is served at `http://<mac-ip>:<http-port>`, and the WebSocket endpoint is at `ws://<mac-ip>:<ws-port>`. Authentication requires a 64-character token extracted from the Director page's HTML. Clients send JSON messages like `{"type":"auth","text":"<token>"}` within five seconds of connecting. Commands include `start` (replaces text and starts word tracking), `updateText` (edits unread text while preserving read position), and `stop`. The server broadcasts state frames every ~100 ms with fields like `words`, `highlightedCharCount`, `totalCharCount`, `fontColor`, `cueColor`, and `audioLevels`. A minimal Python client example is provided in the README. Building from source requires Xcode 26 or later, with the macOS target using Swift 5 language mode. For iOS, you need the iOS 26 SDK and simulator runtime, plus a physical device to verify camera, microphone, recording, and Photos saving. Clone the repository, open `Textream/Textream.xcodeproj`, select the appropriate scheme (Textream for macOS, TextreamiOS for iOS), and run with ⌘R. For iOS, you may need to change the bundle identifier if `dev.fka.textream` is unavailable, and enable Developer Mode on the device. The app also registers a `textream://read?text=...` URL scheme and a macOS Service, allowing text selection in any app to be sent to Textream. The original idea is credited to Semih Kışlar, and the app is made by Fatih Kadir Akin. Privacy and support pages are available at textream.net.