What is AssisT?

Featured on the Chrome Web Store — recognised by Google for quality, user experience, and design best practices. The badge was awarded at the 0.9.1 launch on Global Accessibility Awareness Day, 21 May 2026.

AssisT is a comprehensive Chrome extension that augments your learning abilities, not just assists them. Works inside Canvas, Moodle, Google Classroom and Google Docs, and on any site you choose. We believe technology should superpower every learner—leveling the playing field so everyone can excel.

Whether you’re neurodivergent, neurotypical, a visual learner, an auditory learner, or just want to learn more efficiently, AssisT provides 29 supports grouped into Reading, Writing, Seeing, Focus and Study, plus 10 AI features with four AI modes. Everything except the AI features runs in your browser with no account, no subscription and no API key.

Install AssisT from the Chrome Web Store

How-To Videos

New to AssisT? Watch our short how-to video guides — one for every feature, with filters for dyslexia, ADHD, visual impairment, and more.

Browse all how-to videos →

Enable on All Websites

AssisT uses a privacy-first permission model. By default, the extension works immediately on educational platforms:

Platform	Domains
Canvas LMS	.instructure.com, .canvas.com
Moodle	.moodle.org, .moodlecloud.com
Google Classroom	classroom.google.com
Google Docs	docs.google.com

Want AssisT on Every Website?

You can enable AssisT on all websites with a single click. This is optional and only needs to be done once:

Click the AssisT extension icon in your browser toolbar
Look for the “Enable Everywhere” button in the popup
Chrome will show a permission dialog asking to “Access your data on all websites”
Click Allow to grant permission
Reload any open tabs you want to use AssisT on

That’s it! AssisT now works on every website you visit. You’re in control—you can revoke this permission anytime in Chrome’s extension settings.

Why does Chrome show this message? Chrome requires extensions to explicitly request permission before accessing websites. This protects your privacy by ensuring you consent to where extensions can run. AssisT never collects or transmits your data—everything runs locally.

Quick Start Profiles

Get started instantly with pre-configured profiles in Settings → Preferences → Profiles:

ADHD Focus: Pomodoro timer enabled, progress tracking, simplified interface
Dyslexia Support: OpenDyslexic font, wide spacing, reading progress bar
Sensory Sensitive: No animations, muted colours, media blocking

The 29 Supports

The popup groups every support the same way the extension does. Mind Map is the only one marked beta.

Group	Supports
Reading (9)	Read Aloud, Highlight While Reading, Word by Word, Reading Mode, Dyslexia Mode, Read from Images (OCR), Maths Support, Translate Page, Show Text Stats
Writing (4)	Speech-to-Text, Sticky Notes, Citations, Quick Actions Menu
Seeing (6)	Text Customisation, Reading Guide, Screen Color Overlay, Custom Cursor, Magnifying Lens, Reduced Motion
Focus (5)	Focus Mode, Simplified Interface, Reading Progress, Pomodoro Timer, Auto-play Blocking
Study (5)	AI Assist, Mind Map (beta), Canvas LMS, Moodle LMS, Google Classroom

Beyond Accessibility: Augmentative Technology

Accessibility tools help you cope. Augmentative tools help you excel.

AssisT represents a paradigm shift in educational technology:

Not a crutch, but a superpower — Features like AI summarisation, text-to-speech, and OCR don’t just make content “accessible,” they enhance how everyone learns
Leveling the playing field — Everyone has different learning strengths. AssisT lets you leverage YOUR strengths instead of struggling with your weaknesses
For all learners — Yes, AssisT was designed with neurodivergent students in mind, but neurotypical students benefit just as much. Speed readers use TTS. Honor students use AI summarisation. Visual learners use highlighting. Everyone learns differently.

Think of AssisT as augmented learning — just as augmented reality enhances what you see, AssisT enhances how you learn, comprehend, and retain information.

Key Features

Reading & Listening

Text-to-Speech with Synchronised Highlighting

Listen to any web content using your browser’s built-in voices. Words are highlighted as they’re read, helping you follow along. Customise speech rate (0.5x–2.0x), pitch, volume, and choose from 8 highlight colours.

Keyboard shortcut: Ctrl+Shift+Space (play/pause), Ctrl+Shift+S (stop)

Reading Mode

Distraction-free view that removes navigation, sidebars, and ads. Powered by Mozilla Readability for clean content extraction. Full TTS integration included.

Keyboard shortcut: Alt+R (toggle), Escape (exit)

Dyslexia-Optimised Reading Modes

Three ways of reshaping text, one active at a time:

Syllable colours (the default): alternating colour backgrounds break words into visual chunks
Grammar colours: colours parts of speech (nouns, verbs, adjectives)
Bionic reading: bolds the opening letters of words to guide eye movement

Keyboard shortcut: Alt+Y (toggle)

OCR & Screenshot Tool

Capture and extract text from images, PDFs, and scanned documents. Uses Tesseract.js for offline processing, with 14 recognition languages to choose from. Adjustable upscaling and confidence thresholds, plus an optional accuracy boost that preprocesses the image (contrast, binarisation, sharpening, upscale) before recognition.

Keyboard shortcut: Alt+O (capture)

Maths Support

Numeracy accommodations for dyscalculic and neurodivergent students, off by default. An on-page calculator with a persistent scratchpad that you can operate from the keyboard, and large-number formatting that groups digits, keeps the value copy-safe, speaks the number in words to a screen reader instead of digit-by-digit, and can colour place values. Both tools load only when you switch them on.

Writing & Voice Input

Speech-to-Text

Dictate text directly into any input field using the Web Speech API. Works with standard text inputs, contenteditable elements, and Google Docs rich editor. Features include:

80 spoken command forms for editing, formatting and navigation, with several ways to say each action
Smart auto-punctuation with three modes: Automatic, Assisted, and Manual
Confidence feedback with colour-coded accuracy indicators
Custom vocabulary with domain presets for Medical, Legal, Academic, and STEM
Always-on field detection that automatically shows the microphone button when you focus any text field

Keyboard shortcut: Ctrl+Shift+P (switch STT profile)

Sticky Notes & Annotations

Create floating notes on any webpage. Tag and categorise notes, export them, and view all annotations in a dedicated panel.

Keyboard shortcut: Alt+N (create note)

Study & Research Tools

Dictionary Lookup

Instant definitions for selected words with example sentences and part-of-speech information. Uses the Free Dictionary API.

Keyboard shortcut: Alt+Shift+D

Translation

Translate content into 14 languages: English, Spanish, French, German, Italian, Portuguese, Dutch, Polish, Russian, Chinese, Japanese, Korean, Arabic and Hindi.

Translation providers:

MyMemory (default): Free, no API key required
DeepL (optional): Your own DeepL API key, free or pro plan
Azure Translator (optional): Your own Azure Cognitive Services key

Keyboard shortcut: Alt+T

Mind Map (Beta)

An outline-first way to plan an essay or revise a topic. The outline pane is a WAI-ARIA tree with a full keyboard map, and the canvas is a second view of the same map with pan, zoom and drag re-parenting, so both edit one model.

Six templates: essay plan, lecture notes, compare and contrast, project plan, reading notes, blank
A library of your maps with search, rename, delete and import
File menu: New map, Open, Save, Save as, Import a file, and export to Markdown, OPML, JSON, SVG or PNG. Markdown, OPML and JSON round-trip losslessly
Read aloud speaks a branch with your stored voice and highlights the node being read
Study mode hides children for recall practice, and each miss can be turned into a flashcard
“Add to mind map” in the highlight menu sends a phrase from any page into the map you choose
With an AI provider set up, the AI panel can build a map from the page or from pasted text, suggest sub-points, explain a node into its notes, and turn a branch into recall questions

Mind Map works with AI switched off. “Map this page” turns a page’s headings into branches without asking any model.

Keyboard shortcut: Alt+Shift+M (open Mind Maps)

Read the Mind Map guide →

Citation Management

Capture a source from any page and format it in APA 7th, MLA 9th, Harvard (Cite Them Right) or Chicago 17th (author–date). Your chosen style is remembered.

In-text citations with page locators, copied in a rich and a plain flavour at once
A checker that compares the citations in your essay against your reference list and reports what is in one and not the other, on your device
Quote, locator and note capture when you save a source
Add a source by DOI, and import an existing library from Zotero as BibTeX or RIS
10 source types including book chapters and theses, and sources with no date
Bibliography manager with projects, bulk select, copy and export, and a trash you can restore from
CRAAP test evaluation for source quality
Export to BibTeX, RIS, JSON and CSV

Keyboard shortcut: Alt+C (save the selected text as a citation)

Text Statistics

Track word count, character count, sentence count, paragraph count, and estimated reading time.

Keyboard shortcut: Ctrl+Shift+W

Quick-access toolbar appears when you select text. Default actions include: Read Aloud, Dictionary, Translate, Annotate, Save as Citation, Add to mind map, and Copy. Nine AI actions appear when AI is enabled: Summarise, Simplify, Breakdown, Socratic Tutor, Knowledge Graph, Citation Analyzer, Multi-Document Compare, Study Path and Speed Read. Auto-hides after a delay you can set.

Keyboard shortcut: Ctrl+Shift+H (toggle)

Visual Comfort & Accessibility

Focus Mode

Dims everything outside a clear reading window that follows your cursor. Set the window width (150px to 800px), its height (50px to 250px) and how dark the rest of the page goes (10% to 100%).

Keyboard shortcut: Alt+F

Reading Guide

Horizontal line that follows your cursor to help you keep your place. Choose the colour (Black, White, Red, Blue, Green, Magenta or Orange), the thickness (1px to 10px) and the opacity.

Keyboard shortcut: Alt+G

Screen Color Overlay

Apply a colour tint for visual comfort: Warm Sepia (the default), Soft Yellow, Peach, Cream, Cool Blue, Soft Green, Light Pink or Lavender, at 10% to 90% intensity. Works inside Google Docs and Google Sheets from 1.0.0 onwards.

Keyboard shortcut: Alt+Shift+O

Reading Progress Bar

Visual indicator showing scroll progress. Position it at the top or bottom of the page, in Green, Blue, Orange, Purple or Pink, with the percentage shown if you want it.

Text Customisation

5 Fonts: System default, Lexend (the default), OpenDyslexic, Comic Sans MS, Arial
Spacing Controls: line spacing (1.0-3.0), letter spacing (0-0.5em, 0.12em by default), word spacing (0-0.5em, 0.16em by default), paragraph spacing (1.0-4.0em, 2.0em by default)

All spacing options are WCAG 2.2 SC 1.4.12 compliant.

Reduced Motion

Disables all animations for users with vestibular disorders. Respects the prefers-reduced-motion system preference. WCAG 2.1 SC 2.3.3 compliant.

Custom Cursor & Magnifying Lens

Enhanced cursor visibility in four styles: Crosshair, Circle, Dot with Ring and Large Arrow, sized 24px to 96px, with accurate hotspot positioning for each. The Magnifying Lens is a floating window at 1.5x to 4x that follows your cursor, and can be offset so it does not sit on top of what you are reading. The custom cursor works inside Google Docs and Google Sheets from 1.0.0 onwards.

Productivity

Pomodoro Timer

Configurable work intervals (5 to 60 minutes), short breaks (1 to 30 minutes) and long breaks (5 to 45 minutes), in any corner of the page, with optional auto-start on breaks and a sound when a session ends. Designed with ADHD support in mind.

Media Control

Block autoplay video and audio to prevent distractions and sensory overload.

Neurodivergent User Profiles

Built-in Profiles (12)

Twelve profiles ship with the extension. Apply one in Settings → Preferences → Profiles, then change anything you like afterwards.

Default
Reading Mode
Quiz Mode
Low Vision
ADHD Focus
Autism Comfort
Dyslexia Support
Sensory Sensitive
Night Study
Anxiety Calm
Motor Impairment
Low Vision STT

Speech-to-Text Profiles (7)

Separate speech-to-text configurations, with their own timing, appearance and feedback:

Default: Standard settings
ADHD Focus: Fast response (800ms), minimal distractions, large buttons
Dyslexia Support: Extended processing time (3s), spelling assistance, 3 word alternatives
Anxiety Calm: Forgiving timing (4s), calming colours, gentle feedback
Motor Impairment: Voice activation, extra-large buttons (80px), 60s speech timeout
Low Vision: Maximum contrast, 96px buttons, full audio feedback
Autism Comfort: Predictable behaviour, literal commands, no surprises

AI Features (Optional)

AssisT offers four AI modes to match your privacy and performance needs. Switch between modes instantly using the inline AI mode switcher in the popup—no wizards, no page reloads.

1. Off Mode

No AI processing — all 29 supports still work exactly as they do with AI on. Choose this if you don’t need AI assistance or want to minimise resource usage.

2. Local AI (Ollama)

Privacy: 100% private, runs on your computer
Cost: Free
Requirements: Ollama installation
Best For: Users prioritising privacy and offline access, and institutions that need no text to leave the device

Live model selector — AssisT auto-detects your installed Ollama models and shows them in a dropdown. Switch models without restarting.

Recommended models:

qwen3:8b-q4_K_M (5GB) - Best JSON compliance, recommended default
llama3.1:8b (5GB) - Strong general-purpose performance
gemma3:4b (3GB) - Fast responses for basic tasks
llava (4.7GB) - Image understanding (vision)

Permission - AssisT does not ask for access to your local AI server when you install it. The first time you switch Local AI on, or apply a custom server address, Chrome shows an “Allow” prompt for the loopback address (localhost or 127.0.0.1 on port 11434). Choose Allow and the extension can reach Ollama; until you do, the popup says local AI needs your permission rather than telling you to start Ollama. A student who never uses local AI is never asked. You can withdraw the permission at any time in Chrome’s extension settings.

Server address - the Local AI panel accepts a custom Ollama address as well as the default loopback one, so an install on a non-standard port works without editing anything. The public build only permits loopback addresses; an institution running a shared Ollama server adds its own host to its build.

Reasoning models - qwen3 and deepseek-r1 are handled correctly: their thinking text is kept out of the answer, and JSON-mode requests still get valid JSON back.

3. Browser AI (WebLLM)

Privacy: 100% private, runs in your browser tab
Cost: Free
Requirements: WebGPU-capable browser (Chrome 113+)
Best For: Zero-install AI — no software to install at all

Browser-native AI powered by WebLLM and WebGPU. Models download once and run entirely in your browser—no server, no API, no installation.

Performance note: Browser AI uses WebGPU, which is significantly slower than a native GPU runtime. Responses typically take 30–90 seconds. If you need faster local AI, use Local AI (Ollama) — it runs the same models at full GPU speed.

Available models:

llama-3.2-1b (1GB) - Ultra-fast, basic tasks
gemma-2b (2GB) - Balanced small model
phi-3.5-mini (3GB) - Strong reasoning for size
qwen2.5-3b (3GB) - Good multilingual support
llama-3.2-3b (3GB) - Capable general model
mistral-7b (5GB) - Complex analysis
llama-3.1-8b (6GB) - Most capable browser model

The AI Setup Wizard shows which models are already cached on your device, which need downloading, and lets you remove all downloaded models to free up browser storage.

4. Cloud AI (Your API Key)

Privacy: Data sent to your chosen provider (not Fiavaion)
Cost: Free with Google Gemini; paid for other providers
Providers: Google (Gemini), Anthropic (Claude), OpenAI (ChatGPT), Perplexity
Best For: Maximum capability, latest AI features, complex analysis

Supported providers:

Google Gemini - free API key, no credit card required. Gemini 2.5 Flash is the default
Anthropic - Claude, paid
OpenAI - paid
Perplexity - Sonar, paid

Each provider’s model picker lists the current models for that provider, so the choices follow the provider rather than the extension’s release cycle. Model names are deliberately not printed here: they change more often than this page does.

API keys are encrypted with AES-256 and stored locally—never sent to Fiavaion servers.

Learn more about AI modes and setup →

Inline AI Mode Switcher

Switch AI modes directly from the popup without navigating to settings:

Mode chip row — Click Off, Cloud, Browser AI, or Local AI to switch instantly
Contextual panels — Each mode shows relevant controls inline:
- Cloud: Provider selector, API key input, model dropdown
- Browser AI: Model picker with download/load buttons and progress bars
- Local AI: Live Ollama status with model dropdown
Persistent status bars — Every AI feature shows its current status (available, unavailable, or processing) directly in the popup

8 AI Features

With any AI mode enabled (except Off), you get:

Smart Summarisation: Brief, detailed, bullet points, or ELI5 mode
Text Simplification: Three complexity levels for better comprehension — automatically strips trailing AI conversational artifacts
Assignment Breakdown: Convert instructions to step-by-step checklists
Socratic Tutor: Guided questions for deeper learning — smart hint filtering suppresses generic canned responses
Multi-Document Compare: Analyse and contrast multiple text sources side-by-side — includes Read All TTS button
Knowledge Graph: Visualise concept connections with interactive node graphs — optimised for local AI with smart token limits and person-detection heuristics
Citation Analyzer: Evaluate source credibility using the CRAAP test
Study Path Generator: Create personalised learning roadmaps — includes Copy and Read All buttons

All 8 features use a shared AI routing client that consistently handles mode detection, availability checking, and response generation across all modes.

Turn on “Try the new AssisT” in settings to switch the popup to flat navigation: a bar of group buttons across the top (Reading, Writing, Seeing, Focus, Study), one group of supports on screen at a time, a compact header, and a light or dark theme.

This is a look-and-layout change only. It is not a feature gate: every support, control and button works the same on either popup, and you can switch back at any time.

Companion Features (Preview)

A set of newer supports ships in 1.0.0 under Companion (preview) in the popup. They are on by default and each has its own switch, so you can turn off any one of them without affecting the rest. They are marked preview because they are newer than the rest of the extension and are still being refined.

Study tools discovery - cards in the popup and a first-run tip that point out the AI study tools you get by selecting text
Personalised AI - your learning profile shapes the output of every AI feature, not just two of them
Equation read-aloud - equations join the synchronised-highlight reading flow instead of being skipped
Stuck-reader rescue - a quiet offer of help when the same passage keeps being reread
Authorship Shield - a private writing history kept on your device that you can export as a report if your work is ever questioned
Flashcards - spaced-repetition cards made from your own highlights and from the points you miss in mind map study mode
Summary reading level - a dial that sets how simply AI summaries and simplifications are written
Lecture companion - live captions and structured notes from your microphone
Voice commands - hands-free control, such as “read this page”, “save citation”, “open my library”

For Institutions

A college or university can apply its own name, logo, accent colours and default LMS domain to AssisT with an Ed25519-signed branding key, delivered through enterprise policy (chrome.storage.managed) or pasted into Advanced Settings. Keys are verified offline against a public key bundled in the extension and fail closed if they do not verify. The accent colour is darkened automatically so that contrast still meets WCAG 2.2 AA on every surface. The public build carries no key and is unbranded.

Keyboard Shortcuts

All 16 shortcuts are customisable in settings. Chrome-reserved shortcuts are blocked automatically, and every shortcut needs at least one modifier key, with Escape as the one exception.

Action	Shortcut
TTS: Play/Pause	`Ctrl+Shift+Space`
TTS: Stop	`Ctrl+Shift+S`
OCR: Activate Screenshot	`Alt+O`
Reading Mode: Toggle	`Alt+R`
Reading Mode: Exit	`Escape`
Dictionary: Lookup Selected Text	`Alt+Shift+D`
Text Statistics: Toggle Display	`Ctrl+Shift+W`
Highlight Menu: Toggle	`Ctrl+Shift+H`
Sticky Notes: Create New Note	`Alt+N`
Translation: Toggle	`Alt+T`
Citations: Save Selected Text	`Alt+C`
Open Mind Maps	`Alt+Shift+M`
Focus Mode: Toggle	`Alt+F`
Reading Guide: Toggle	`Alt+G`
Screen Overlay: Toggle	`Alt+Shift+O`
Dyslexia Mode: Toggle	`Alt+Y`

Speech-to-Text has its own profile-switch key, Ctrl+Shift+P, set in the STT profile settings rather than in this list.

If you saved a custom shortcut set before this release, it picks up any new default keys rather than ignoring them, so Alt+Shift+M and Alt+C are there without you rebuilding the set.

Privacy & Data

AssisT is built with privacy as a core principle:

No tracking: We don’t collect any user data
No accounts: Nothing to sign up for, nothing to sign in to
Local by default: Your text, documents and images stay on your device unless you switch on a cloud AI provider or a cloud translation provider yourself
Local settings only: Your preferences are stored in chrome.storage.local, on the device you set them on. Nothing syncs across devices. Annotations, citations, mind maps and the Authorship Shield writing history are stored on your device in IndexedDB and are never uploaded
GDPR: Fiavaion processes no personal data. Any cloud AI or translation provider you enable is your own sub-processor, under your agreement with them
FERPA: designed with FERPA principles in mind — no student data is collected

Translation uses MyMemory by default (free, no API key). DeepL and Azure Translator are optional and need your own API key. Whichever provider you use, the text you select is sent to that provider; Fiavaion never sees it.

Read our full Privacy Policy →

Accessibility Standards

WCAG 2.2 Level AA is the standard AssisT is built and tested against. The Accessibility Conformance Report for 1.0.0 is a self-assessment by the developer, not an independent audit: it records 38 of 42 applicable criteria as fully supported and 4 as partially supported, with none unsupported, and it flags the 1.0.0 features that have not yet been tested against those criteria. Screen-reader testing with NVDA, JAWS and VoiceOver has not been done, and a third-party audit is planned but not commissioned. What is in place today:

Semantic HTML structure throughout
Full keyboard navigation support
Screen reader compatible with ARIA labels
High contrast mode available
Focus indicators on all interactive elements
Text spacing meets SC 1.4.12 requirements
Reduced motion respects SC 2.3.3

What’s New

Release history is published on GitHub Releases and rendered automatically below. The full versioned CHANGELOG is in the repository.

Latest - v1.0.0 - First stable release

Everything since 0.9.3: Mind Map (beta), Maths Support, institutional branding keys, a configurable Ollama address, the OCR accuracy boost, the new popup, the rebuilt citation manager and the companion preview features. The full notes are in the changelog.

v0.9.3 - 29 May 2026

Citation manager overhaul, cross-browser API abstraction, the lms-adapter and local-ai-router package scaffolds, a pluggable prompt registry, and a Firefox build target.

v0.9.2 - 23 May 2026

Ollama model catalogue — 11 curated models by category (General, Reasoning, Multilingual, Code, Vision) with hardware tiers (4 GB / 8 GB RAM)
“Remove all downloaded models” button in Browser AI setup — clears WebLLM cached model storage
Browser AI speed warning added to setup wizard — honest 30–90s caveat; directs speed-sensitive users to Ollama

v0.9.1 — 21 May 2026 — GAAD Launch

First public Chrome Web Store release, launched on Global Accessibility Awareness Day 2026. Awarded the Chrome Web Store Featured badge by Google.

v0.9.0 — 15 May 2026 — Final Beta

EUPL-1.2 licence, VPAT 2.5, GDPR font bundling, 27 MB → 11 MB bundle, XSS/security audit, TTS Reading Scope + Word Highlight Speed, full gap audit.

View the full release history on GitHub →

Technical Details

29 supports across Reading, Writing, Seeing, Focus and Study, plus 10 AI features
4 AI modes: Off, Cloud, Browser AI (WebLLM), Local AI (Ollama)
16 customisable keyboard shortcuts
80 spoken command forms while dictating, plus a separate hands-free voice palette
12 built-in profiles in the popup, plus 7 speech-to-text profiles
5 fonts: System default, Lexend, OpenDyslexic, Comic Sans MS, Arial
4 citation styles: Harvard (Cite Them Right, the default), APA 7th, MLA 9th, Chicago 17th, across 10 source types
14 languages for translation
14 recognition languages for OCR text extraction
8 WebLLM models available for browser-native AI
Licence: EUPL-1.2 — free to use, modify and redistribute under the same licence. The source is public

For learning support and disability services

If you are assessing AssisT for a college or university, everything a disability support officer needs to make an adoption decision is collected on one page: the Accessibility Conformance Report, the privacy and data-processing statement, the sub-processor list, the managed deployment guide, the licence and the support commitments.

AssisT for learning support and disability services →

Support Development

Love AssisT? Help us continue building accessible learning tools for everyone. Look for the Support button in the extension popup.

Support the Project →

AssisT