Aletheiaby Lonia AI

Features

Every barrier has an answer.

Aletheia is built around the most common ways digital content fails people with disabilities: missing alt text, image-only PDFs, broken page structure, and forms with no labels. Each feature is a direct response to a specific failure, and each runs on the user's own device by default.

Core capabilities

Read the page, then rebuild it

Page understanding

Aletheia parses a web page or document and identifies the barriers a screen reader hits: images without meaningful alt text, image-only PDFs, missing or out-of-order headings, absent landmarks, and unlabeled form fields. Detection happens on your device.

Restructuring and sanitization

Content with no headings, no landmarks, and no logical reading order is rebuilt into a properly structured document. Headings are inferred from the visual hierarchy, reading order is corrected, landmarks are added, and clutter that gets in a screen reader's way is removed. The accessible HTML export rebuilds this structure.

OCR for scanned PDFs

Image-only PDFs are the most common inaccessible document in education and government. Aletheia runs OCR on your device to turn a picture of text into real, searchable, screen-reader-navigable content. Works offline after one-time setup download.

Form labeling

Forms with missing labels, broken tab order, and unlabeled inputs are scanned, and Aletheia identifies the unlabeled fields so a screen reader user can tell where input is required.

Local storage library

Every accessible version you create is saved on your device, organized and searchable, and available without rescanning. Export it or clear it at any time. It is yours.

Accessible export

Save the rebuilt content as accessible HTML, ready to open in the reader or assistive technology you already use. Export to tagged accessible PDF. Latin-script content (English, Spanish, French, German, and other European languages) exports cleanly with proper structure tags, alt text, and outlines. Non-Latin scripts may show fallback characters in the current build. Export to audio. Uses an on-device text-to-speech model (about 90 MB, downloaded once and cached). All processing stays on your device. Export to plain text. Unicode content (including non-Latin scripts) passes through losslessly.

AI description architecture

Three ways to describe an image, and you choose

When an image has no useful alt text, Aletheia can describe what it shows. How that description is produced depends on your tier, and the default everywhere is on-device.

On-device descriptions

The default on every tier, including Free. A compact vision model runs in your browser to describe images with no network call, no per-image cost, and no data leaving the device. Works offline after one-time setup download. This is the only AI path on the Free, Student, and Campus tiers.

Enhanced AI, opt-in

On Personal and Family, you can opt in to send a single image to Lonia-provided cloud AI for a higher-detail description. Personal includes up to 500 Enhanced descriptions per month. Family includes up to 1,500 per month, shared across up to five users. It is off until you turn it on.

Bring your own key

Professional and Enterprise connect your own Anthropic, OpenAI, or OpenRouter key for cloud-grade descriptions at your own cost, with on-device fallback when no key is set. Aletheia does not resell cloud AI on these tiers. Your API key is stored encrypted on your device using AES-GCM-256 with a non-extractable key. It never leaves your browser and is never sent to Lonia.

For professionals

Built for billable remediation work

Document mode

Professional adds full document remediation: process an entire document end to end, not one element at a time, and export a complete accessible version. Pair it with your own AI key for cloud-grade descriptions across a whole file.

Activity log

Professional records an activity log of what was processed and when, so accessibility consultants and remediation specialists have a record of remediation actions for billable client work.

For families and students

Plans shaped around how people actually use them

Family seat management

Up to five users, each on their own device with their own private library, under a single subscription and one bill. The Enhanced AI allowance of up to 1,500 descriptions per month is shared across the household.

Student LMS helpers

For verified students with a .edu email. Aletheia provides course organization and assignment tracking, keeping accessible versions sorted by course so a semester of readings stays in order.

Privacy and security

The safest data policy is not holding the data

On-device processing

Documents are parsed, restructured, and stored in your browser. They are not uploaded to a server and never leave your device unless you export them yourself or opt in to send a single image for Enhanced AI.

No central document store

Aletheia keeps no server-side database of your documents. There is nothing in the middle to breach, subpoena, or sell, because the documents are not on a server.

OAuth sign-in only

Sign in with Google or Microsoft. Aletheia stores no passwords and offers no password login. The Free tier requires a Google or Microsoft account; no credit card is needed.

No tracking, no telemetry

No third-party analytics, advertising trackers, or telemetry. Account data is limited to what is needed to apply your plan and process billing through Stripe.

Accessibility

The tool that fixes accessibility is itself accessible

If a tool that promises accessibility is itself inaccessible, the promise is broken before the work begins. Aletheia's interface meets the standard the tool helps people enforce.

WCAG 2.2 AA and Section 508

Contrast ratios, focus indicators, target sizes, and structure are tested against WCAG 2.2 AA and aligned with Section 508 across every screen.

Keyboard and screen reader

Every action is reachable with Tab, Enter, and Escape, with no mouse-only controls. Tested with NVDA, JAWS, and VoiceOver, using semantic HTML and ARIA only where needed.

Motion, contrast, and touch

Animation respects prefers-reduced-motion, a high contrast presentation is available, and mobile touch targets meet a 44 by 44 pixel minimum.

Platform support

Where Aletheia runs

Android Chrome

The primary mobile target. Aletheia runs in Chrome on Android with full keyboard, screen reader, and touch support.

iOS Safari

Aletheia runs in Safari on iPhone and iPad, with VoiceOver support.

Desktop browsers

Aletheia runs in current desktop browsers for users who do their reading and remediation on a larger screen.

Start with the features that ship today.