Complete Guide

Everything you need to know to use SacredTexts AI and discover the connections among the sacred texts of the great religious traditions.

Version 2.21.0

Overview

SacredTexts AI is a progressive web application (PWA) that leverages artificial intelligence to analyse and visualise the connections among the sacred texts of the world's major religious traditions.

The application combines modern machine-learning technologies with interactive visualisations and data sonification to provide a unique tool for the comparative study of religions, accessible to researchers and enthusiasts alike.

Advanced AI Analysis

Uses Google Gemini to extract themes, concepts and cross-references from sacred texts.

Graph Visualisation

Interactive representation of concept connections with physics-based simulation.

Export

Export to JSON-LD and RDF/Turtle for interoperability with research tools.

Installable PWA

Installable as a native app on any device, with offline support for browsing history.

Accessibility

Designed with accessibility in mind, including screen-reader support and keyboard navigation.

Getting Started

Initial Setup

Obtain a Gemini API Key

Go to Google AI Studio, sign in with your Google account and create a new free API key.

Configure the Application

Click the gear icon (⚙️) in the top bar, paste your API key into the designated field and save the settings.

Select a Text

Choose a sample text from the dropdown menu or paste a passage from any sacred text directly into the text field.

Run the Analysis

Click the "Analyse with AI" button and wait for the artificial intelligence to process the text. The results will appear in the graph and in the side panel.

Enable Audio (Optional)

Click the volume button in the graph header to activate sonification. Use the preset selector to choose among 5 generative audio presets (1-5).

💡 Tip

Install the application as a PWA for a better experience! On Chrome, click the install icon in the address bar.

Features

Main Interface

The interface is divided into three main panels:

Navigation Bar

Graph Controls

History and Favourites

Every analysis is automatically saved in the local history. You can:

Connection Graph

The graph is an interactive visual representation of the semantic structure of the analysed text. It uses a physics-based simulation algorithm to position nodes organically.

Hierarchical Structure

The graph features three levels of nodes:

  1. Central Node (Source): Represents the analysed text. It is the focal point from which all themes branch out.
  2. Thematic Clusters: Groupings of related concepts, colour-coded by category (theology, ethics, mysticism, practice).
  3. Child Nodes (Children): Specific concepts extracted from the text, linked to their parent cluster.

Interactions

Action How to
Zoom Mouse wheel or + / - buttons
Pan (move view) Click and drag on the background
Move a node Click on the node and drag
View node details Quick click on the node (without dragging)
Hover info Hover over a node to see the tooltip
Reset view Reset button in the toolbar

Node Details Sidebar

Clicking on a node opens a sidebar that displays:

Colour Coding

Theology

Concepts relating to the divine nature, doctrines and fundamental beliefs.

Ethics

Moral precepts, rules of conduct and justice.

Mysticism

Spiritual experiences, contemplation and union with the divine.

Practice

Rituals, prayers, daily practices and observances.

Further Reading

For a detailed explanation of how the AI prompt, the validation pipeline and the graph's physics parameters work, see the Behind the Scenes page.

🎵 Audio Sonification

SacredTexts AI features a generative audio engine (GenerativeAudioEngine in generative-audio.js) powered by Tone.js. The engine is fully synchronized to the graph visualization: nodes trigger notes as they appear, and user interactions shape the sound in real time. Audio is disabled by default.

How to Enable Audio

  1. Click the volume button (🔊) in the graph header
  2. On the first click, the browser will request audio permission
  3. The button turns gold when audio is active
  4. Use the preset selector to choose among 5 generative audio presets

The 5 Audio Presets

1 (default)

Warm PolySynth with long release. Slow tempo, high reverb, gentle FM drone. Ideal for meditative exploration of the graph.

2

Bright FMSynth with short attack and moderate release. Crisp bell-like tones with medium reverb and subtle delay. Clear, sparkling textures.

3

Deep AMSynth in a low register. Slow tempo with heavy reverb and prominent drone. Evokes ancient, ritualistic soundscapes.

4

Warm FMSynth with vibrato. Mid register, low drone, dark filter. Enveloping and meditative timbre.

5

Percussive PolySynth with short release. Faster tempo, dry mix with minimal reverb. Energetic pulse synchronized to node activity.

Tradition-Specific Musical Scales

The generative engine automatically selects a musical scale matching the detected religious tradition. Every note produced — whether from node interactions or the background generative layer — stays within the tradition's authentic scale:

Tradition Scale Notes
✝️ Christianity Ionian mode (Gregorian) C D E F G A B
✡️ Judaism Ahava Rabbah C Db E F G Ab B
☪️ Islam Maqam Hijaz C Db E F G Ab Bb
🕉 Hinduism Raga Bhairav C Db E F# G Ab B
☸️ Buddhism Pentatonic C D F G A
☯️ Taoism Chinese pentatonic (Gong) C D E G A
General / Other Natural Minor C D Eb F G Ab Bb

Each tradition also applies specific parameter modifiers to the active preset: register (octave offset), tempo multiplier, drone presence and tuning, multi-note probability, reverb wet level, and filter cutoff frequency.

Node-Audio Interaction

Interactions with graph nodes produce distinct audio responses:

Audio Info Bar

When audio is active, an info bar appears below the graph header displaying: the current scale name, the notes in the scale, and the active synth type (PolySynth, FMSynth, AMSynth). This provides a transparent view of the generative parameters in use.

Optional FM Drone

Each preset can include an FM drone — a sustained low-frequency tone that provides harmonic grounding. The drone's presence, volume and frequency are configured per preset and further adjusted by tradition-specific modifiers. Presets 1 and 3 feature a prominent drone; preset 5 omits it entirely.

Technology

The generative audio engine is built on Tone.js and uses PolySynth, FMSynth and AMSynth with configurable effects chains (reverb, delay, filter, stereo panner). No pre-recorded audio samples are required — all sound is synthesized in real time.

💡 Tip

The engine auto-selects the best preset based on the detected tradition. Try switching presets manually to hear the same graph data rendered through different sonic textures. Drag nodes slowly across the canvas to explore the full range of real-time audio modulation.

Data Export

SacredTexts AI supports exporting analyses in standard formats for interoperability with academic research tools and linked data.

Available Formats

JSON

Native application format. Contains all analysis data in a structured format.

JSON-LD

JSON for Linked Data. Uses Schema.org and Dublin Core vocabulary for maximum interoperability.

RDF/Turtle

RDF format in Turtle syntax. Ideal for import into triple stores and SPARQL endpoints.

How to Export

  1. Run an analysis or load one from history
  2. In the "Detailed Analysis" panel, click the Export button
  3. Select the desired format from the dropdown menu
  4. The file will be downloaded automatically

Semantic Vocabulary

The JSON-LD and RDF formats use the following vocabularies:

Academic Use

The JSON-LD and RDF formats comply with Semantic Web standards and can be imported into tools such as Protege, Apache Jena, or published as Linked Open Data.

Detailed Analysis

The Detailed Analysis panel presents AI processing results in a structured textual format. The analysis engine supports two free user-selectable Google Gemini models: Gemini 2.5 Flash-Lite (default, fast and cost-effective) and 2.5 Flash (most accurate). The model is selected from Settings (≅). Each model has an independent API quota.

Analysis Sections

⚠️ Important Note

The analysis is generated by an AI model and may contain inaccuracies or interpretations not academically validated. It is recommended to verify the information with scholarly sources for research use. The AI may reflect biases present in its training data.

AI Transparency

To view the exact prompt sent to the AI, model comparisons and the text curation methodology, visit Behind the Scenes.

Audio

SacredTexts AI features a single generative audio engine (GenerativeAudioEngine) built on Tone.js. All sound is synthesized in real time — no pre-recorded samples are used. Audio is disabled by default: click the 🔊 button in the graph toolbar to enable it.

Presets

Five selectable presets shape the overall sonic character. Use the preset selector in the graph toolbar to switch between them:

Activation

Node-Audio Interaction

Graph node interactions produce distinct audio responses:

💡 Tip

Try different presets and slowly drag a node from the centre towards a corner to explore the full range of real-time audio modulation. The audio info bar below the graph header shows the active scale, notes, and synth type.

Supported Traditions

SacredTexts AI includes sample texts from eight major world religious traditions. You can also analyse texts from any tradition by entering them manually.

✝️
Christianity
☪️
Islam
✡️
Judaism
🕉️
Hinduism
☸️
Buddhism
☯️
Taoism
🔯
Sufism
☦️
Orthodoxy

Included Sample Texts

⚠️ Disclaimer

The texts in the database are provided for educational and informational purposes. All translations are drawn from recognized academic editions: NRSV (1989) for biblical texts, and established scholarly translations for other traditions (Chadwick/Oxford, Kavanaugh & Rodriguez/ICS, Vannini/Adelphi, Abdel Haleem/Oxford, D.C. Lau/Penguin, etc.). Each text includes full metadata with translator, publisher and publication year. For academic or research use, it is still recommended to verify the texts against the original critical editions (Nestle-Aland, BHS, Sources Chrétiennes, etc.).

Sources and Verification

For detailed information on translation sources, critical reference editions and the AI-assisted curation process, see Behind the Scenes - Text Selection and Translation.

⚙️ Settings

Google Gemini API Key

The API key is required to use the AI analysis. For security reasons, the key is saved only in the current browser session and will need to be re-entered when the application is reopened.

⚠️ Security

Never share your API key. If you suspect it has been compromised, regenerate it immediately from Google AI Studio.

Comparative Study Mode

This option relaxes the AI safety filters to allow the analysis of religious texts that might otherwise be blocked (e.g. controversial passages, apocryphal texts, cross-tradition comparisons).

Note: This mode is intended exclusively for academic and comparative study of religions. The user is responsible for appropriate use.

Automatic Updates

The application automatically checks for new version availability. When an update is detected, a prompt will appear to reload the page and obtain the latest features.

Roadmap

SacredTexts AI is under continuous development. Here is the feature status overview:

Completed in v2.2.0

Completed

Data Sonification

Generative audio engine (Tone.js) with 5 presets, 7 tradition-specific scales, 35 melodic fragments, FM drone and real-time node interaction.

Completed

JSON-LD/RDF Export

Export in Semantic Web standard formats for academic interoperability.

Completed

Node Details Sidebar

Side panel with complete information, navigation and actions on graph nodes.

Completed

Cache Busting System

Automatic updates with new version detection and reload prompt.

Completed in v2.9.16

Completed

"Behind the Scenes" Page

New interactive page with 7 sections: text inventory with critical sources, AI prompt anatomy, graph mechanism, AI model comparison, tools and timeline, prompt playground, credits and methodology.

Completed

Translation Metadata in Texts

Added metadata field to all 48 sacred texts with Italian translation source, critical edition in original language and academic notes.

Completed

Integrated Navigation

New "Behind the Scenes" button in the navbar, link in the Info modal and cross-references in the user guide.

Completed

AI Research Leads

Automatic generation of epistemological questions based on text analysis. Click to copy, double-click to re-analyse.

Completed

AI Confidence (X-Ray Mode)

Visualisation of AI model certainty: opacity proportional to confidence, dashed borders for low values, colour-coded percentage badge.

Completed

Cache and Update Management

Styled update banner, buttons for update checks and cache clearing in settings, complete SKIP_WAITING flow.

Completed

Dynamic Source Header

Automatic identification of the analysed text with title, author, tradition and emoji in the results panel.

Completed

Generative Audio Engine with Node Interaction

Full generative audio engine with 5 presets, tradition-specific scales and melodic fragments, FM drone, node click/hover/drag interaction and audio info bar.

Completed

AI Model Selector

Selector with 2 free Gemini models (2.5 Flash-Lite, 2.5 Flash). Default: Gemini 2.5 Flash-Lite. Each model has an independent API quota.

Completed

Welcome Splash with Version and Update Check

Welcome screen at launch displaying the current version and automatic update availability check.

Completed

429 Retry Fix with Exponential Backoff

Intelligent handling of rate limiting errors (HTTP 429) with an exponential backoff retry strategy to ensure API call stability.

Planned

Planned

AI Disclaimer and Validation

"Not academically validated" badge on results with explanatory tooltip.

Planned

Academic Presentation Mode

Layout optimised for projection with enlarged fonts, simplified controls and slide export.

Completed

Tradition-Specific Generative Audio

7 authentic musical scales (Ionian, Ahava Rabbah, Maqam Hijaz, Raga Bhairav, Pentatonic, Chinese Gong, Natural Minor), 35 melodic fragments per tradition, tradition-specific parameter modifiers (register, tempo, drone, reverb, filter). Automatic tradition detection from dropdown or AI analysis.

Completed

Immersive Graph Animations

Slower and smoother node movements with recalibrated physics (damping 0.92, halved velocity). Camera transitions 1800ms with ease-in-out-quart easing. Slowed node pulsation for a “breathing” effect.

Planned

Textual Variant Comparison

Different translations of the same passage with difference highlighting and philological notes.

Planned

Intertextuality Visualisation

Connections between texts from different traditions with a multi-tradition graph and connection type filters.