WhaleRead

About WhaleRead

WhaleRead is a local-first macOS reader for TXT, Markdown, and EPUB files. It translates content using on-device 7B or self-hosted 30B private models while preserving a bilingual reading layout. The tool keeps your library under local control and includes a human-confirmed review flow alongside an Ask AI feature for page-based questions.

Review

WhaleRead 1.18 is a private beta that runs translation and reading in one macOS workspace. It doesn't send your books to a consumer cloud-text stays on your Mac or on infrastructure you control. The current release handles English and Chinese interfaces, EPUB preservation, and resumable translation, with a review system that surfaces model mistakes rather than hiding them.

Key Features

• Bilingual side-by-side reading for TXT, Markdown, and EPUB files
• On-device 7B model for local translation, with an option to connect a self-hosted 30B model
• Human-confirmed review flow that shows source text, explanation, suggested revision, and final fix
• Page-based Ask AI that works with visible context from the current page
• Resumable translation and EPUB structure preservation

Pricing and Value

WhaleRead is listed as free. The reference content does not specify whether this applies to the current beta period only or to the full 1.18 release. No subscription tiers, one-time purchase prices, or future pricing models are mentioned.

Pros

• Translation happens entirely on-device or on a self-hosted model-your reading library stays under your control
• Review flow exposes model errors directly, showing the original, the flawed translation, and the corrected version
• EPUB files keep their structure during translation instead of being flattened
• Ask AI pulls context from the page you're reading rather than requiring manual copy-paste
• 18 translation directions with native language names in the interface

Cons

• Mac-only-Windows and Linux users can't run the tool
• The 7B on-device model produces literal, awkward translations for idiomatic phrases, as shown in the launch demo
• Not well suited for readers who rely on cloud-synced libraries across multiple devices

WhaleRead fits readers who want translation without handing their library to a third-party service, and who are comfortable running a private beta on Apple silicon. Multilingual readers and translators who want to inspect model output rather than accept it blindly will find the review flow directly useful. Casual readers looking for a polished, cross-platform cloud reader should wait for a more mature release.

Open 'WhaleRead' Website