Home Features Demo Docs Open Source Contact
  • Wiki ⬇ Free Download
  • v1.4.0 · Updated 2026

    AnchorCast Documentation

    AnchorCast is a free, open-source AI-powered desktop app for Windows that listens to your preacher in real time, detects Bible verses, and displays them on screen automatically — with no internet connection required after install.

    🎁

    100% Free, forever. No subscription, no trial, no credit card required. AnchorCast is MIT-licensed open-source software built by and for churches.

    Get StartedGo from install to your first verse on screen in minutes.
    System RequirementsWindows 10/11 · 64-bit · 8GB RAM · GPU optional.
    How It WorksAudio → Whisper AI → Pattern match + semantic search → Display.
    Open SourceMIT license. View, fork, and contribute on GitHub.

    Installation

    Download the latest installer from the download page or from the GitHub releases page.

    1. 1

      Run AnchorCastSetup_v1.4.0_full.exe and follow the installer prompts.

    2. 2

      On first launch, AnchorCast sets up its internal components (Whisper AI model, Bible index, etc.). This takes 1–3 minutes.

    3. 3

      When the main window opens, the interactive tutorial will guide you through the interface. Restart it anytime from Settings → Help → Restart Tutorial.

    ℹ️

    No internet required after install. AnchorCast downloads AI models during setup. After that, everything runs offline on your machine.

    Your First Session

    When AnchorCast opens, you'll see the main dashboard with six panels:

    Live TranscriptTop left — real-time transcription of the sermon as the preacher speaks.
    Program PreviewTop center-left — shows how the verse will look before going live to screen.
    Live OutputTop center-right — what's currently being broadcast to your projector or NDI.
    QueueTop right — verses waiting to be displayed, in order.
    Scripture SearchBottom left — look up any verse by reference or topic.
    AI DetectionsBottom right — verses automatically detected from the live sermon.

    To get started:

    1. 1

      Click Start Transcript in the Live Transcript panel.

    2. 2

      Select your microphone when prompted (or set it in Settings → Microphone beforehand).

    3. 3

      Start speaking or play a sermon — detected verses appear in the AI Detections panel.

    4. 4

      Click the icon on any detected verse to send it to Preview.

    5. 5

      Press L or click Go Live to broadcast it to your display output.

    Transcription

    AnchorCast uses Whisper AI to transcribe sermon audio in real time. Choose between two modes:

    Offline Mode (Recommended)

    Whisper AI runs 100% locally on your PC. No audio ever leaves your machine. Works without internet. Requires a capable CPU or GPU — see System Requirements.

    Online Mode

    For faster transcription on lower-spec machines, audio can be streamed to a cloud speech-to-text service. Requires an internet connection.

    Microphone Setup

    Go to Settings → Microphone to select your audio input device. AnchorCast remembers your selection between sessions.

    💡

    Best practice: Use a USB audio interface connected to your church's mixing board. The improvement in transcription accuracy over a laptop microphone is dramatic.

    Confidence Indicators

    LevelScoreMeaning
    🟢 High80%+Very accurate transcription
    🟡 Medium60–80%Mostly accurate, some words may be off
    🔴 LowBelow 60%Noisy audio or unclear speech

    Verse Detection

    AnchorCast detects Bible verses in two ways, combining both for maximum coverage:

    Direct Reference Detection

    When the speaker says something like "Turn to John 3:16" or "In the book of Psalms, chapter 23", AnchorCast picks it up immediately using pattern matching. Works for standard references, spoken references, abbreviated book names, and context-aware references. Detection is near-instant (under 5ms).

    Semantic Search (AI Detection)

    When the speaker quotes or paraphrases scripture without giving a reference, AnchorCast uses AI to find the matching verse. Spoken text is converted to an embedding and searched against an index of 31,000+ Bible verses. Takes 200–400ms for new queries; common queries are cached for near-instant results.

    Confidence Scores

    TypeScore RangeNotes
    Direct references90–100%Near-certain matches
    Semantic matches35–89%Higher = closer to verse content
    High confidence80%+Auto-queued in Auto display mode
    Below threshold<35%Filtered out, not shown

    The Scripture Search panel lets you look up verses manually. Toggle between two modes with the Tab key.

    Book Mode

    Look up a specific verse when you know the reference. Type a book name, chapter, and verse — AnchorCast autocompletes book names as you type.

    Context Mode

    Find a verse by topic, theme, quote, or paraphrase. Uses the same AI-powered semantic search as automatic detection — searching "God's love for humanity" will find John 3:16 even without those exact words in the verse text.

    Supported Translations

    TranslationCode
    King James VersionKJV
    New King James VersionNKJV
    New International VersionNIV
    New Living TranslationNLT
    English Standard VersionESV
    New American Standard BibleNASB

    Song Manager

    AnchorCast includes a full song library and projection system. Access it from the Songs tab in the main window.

    Display & Output

    AnchorCast can send its verse display to external screens and broadcast software using several methods.

    Auto Mode

    Automatically displays the highest-confidence detected verse. A 2.5-second cooldown prevents rapid flickering. Best for mid-week Bible studies or hands-off operation.

    Manual Mode

    Nothing goes to broadcast until you explicitly send it. Detected verses still appear in the AI Detections panel and queue, but you decide which ones to display and when. Best for Sunday services.

    You can switch between Auto and Manual mode at any time during a session.

    Countdown Timer

    A standalone countdown timer window that can be displayed on a separate monitor or projected to the congregation before the service.

    Start / Pause Timer
    Space
    Reset Timer
    R

    Wi-Fi Remote Control

    AnchorCast runs a local web server on your PC. Open any browser on a phone or tablet on the same Wi-Fi network to control the app remotely.

    1. 1

      Enable Wi-Fi Remote in Settings → Remote Control.

    2. 2

      Note the IP address shown (e.g., 192.168.1.10:8080).

    3. 3

      Open that address in a browser on your phone or tablet.

    The remote lets you advance song slides, send verses to live, control the timer, and toggle On Air status.

    NDI Output

    AnchorCast can send its output as an NDI source to any NDI-compatible software on your network — including OBS Studio, vMix, Resolume, and ProPresenter.

    The NDI source appears as "AnchorCast Main" in your software's source list. Supports alpha transparency, configurable FPS (24/30/60), and resolutions up to 4K.

    ℹ️

    NDI requires both machines to be on the same local network. Install the free NDI Tools package if your broadcast software doesn't detect the source automatically.

    Keyboard Shortcuts

    Go Live / Off Air
    L
    Clear Screen
    C
    Next Song Slide
    Previous Slide
    Start / Stop Transcript
    T
    Open Settings
    Ctrl,
    Bible Search
    CtrlF
    Start / Pause Timer
    Space
    Reset Timer
    R
    Toggle Auto / Manual
    A
    Fullscreen Output
    F11
    Quit
    CtrlQ

    System Requirements

    ComponentMinimumRecommended
    OSWindows 10 64-bitWindows 11 64-bit
    CPUIntel Core i5 / Ryzen 5Intel Core i7 / Ryzen 7+
    RAM8 GB16 GB+
    GPUNot requiredNVIDIA GPU (faster Whisper)
    Storage4 GB free8 GB+ (all Whisper models)
    AudioAny microphoneUSB audio interface
    NetworkNot requiredWi-Fi for remote / NDI

    Troubleshooting

    Transcription is slow or laggy

    Try switching to the tiny or base Whisper model in Settings → Transcription → Model Size. Smaller models use less CPU but are slightly less accurate. If you have an NVIDIA GPU, ensure CUDA is enabled.

    Verse detection is missing references

    Check your microphone levels in Settings → Microphone. Low input gain or background noise significantly impacts detection. Consider using a direct audio feed from your mixer board.

    NDI source not appearing in OBS

    Ensure both machines are on the same local network. Try disabling the Windows Firewall temporarily to test. Install NDI Tools and use NDI Monitor to verify AnchorCast is broadcasting.

    App crashes on startup

    Delete the %AppData%\AnchorCast\models folder and relaunch — this forces AI models to re-download and can fix corrupted installs. If the issue persists, open a GitHub issue.

    FAQ

    Is AnchorCast really free?

    Yes — completely free, forever. No subscription, no trial period, no credit card required. AnchorCast is MIT-licensed open-source software.

    Does it work without internet?

    Yes. After the initial setup (which downloads AI models), AnchorCast runs entirely offline. No audio, no data, nothing leaves your machine in offline mode.

    Does AnchorCast work on Mac or Linux?

    Not officially yet — the current release is Windows only. macOS and Linux builds are planned. Follow progress on GitHub Discussions.

    How accurate is verse detection?

    Very accurate for direct scripture references (90–100% confidence). Semantic detection of paraphrased scripture varies based on audio quality and how closely the speaker's words match the verse text. A clean audio feed from your mixer board makes a significant difference.

    Can I use AnchorCast with ProPresenter or MediaShout?

    Yes, via NDI. AnchorCast sends its output as an NDI source that you can pull into any NDI-compatible presentation software. You can also use HDMI output to a secondary monitor.

    How do I contribute or report a bug?

    Open an issue or pull request on GitHub. We welcome contributions of all kinds — code, documentation, translations, and bug reports.

    ⬇ Download AnchorCast Free View on GitHub ↗