AI NewsProductivityAnnouncement

FluentRead adds one-click bilingual arXiv and PubMed reading in v0.0.41

FluentRead v0.0.41 turns an arXiv or PubMed Central page into a bilingual reader with one click.

AI News

Editorial2 min read

LinkedInX
Social preview card for the FluentRead GitHub repository showing its name, description and star count

Image: FluentRead

Why it mattersResearchers who read papers in English all day can now get a second-language view of each one in place. The paper stays on the same page.

Reading an academic paper in a second language usually means leaving the page. You paste the text into a translator. You lose the figures, the formulas and the links. FluentRead, an open-source browser extension, has a new version that keeps you on the page.

FluentRead v0.0.41 was released on 8 November on GitHub by the project "FluentRead/FluentRead". The repository is ranked 4th on Trendshift's daily list on 10 November, with 9,928 stars in total. It was first published in December 2023. The code is under GPL-3.0.

What is new in this version

The headline change is for academic reading. The release notes say: "Open an arXiv paper in bilingual mode from its abstract page". The same is true for PubMed Central full-text pages. The maintainer, Bistutu, also added shortcuts for Hugging Face Papers, ACL Anthology, CVF Open Access and NeurIPS Proceedings.

The second new feature is for video. The extension can now use a local Qwen3 speech-to-text model for X videos. In-browser Whisper is still the default. The release note adds "subtitle outline and background controls" and preserves labels inside article images when translating PDFs.

A list of bug fixes closes the release. It names Firefox Content Security Policy errors at startup, Safari toolbar popup sizing, and GitHub README hover translation problems.

What the extension does

FluentRead shows a translation next to the original text on a web page. It handles single-word selection, full pages, PDFs, images, and video subtitles. It also supports document formats like ePub and DOCX.

The user picks the translation service. The free ones work with no key. Paid services like DeepL work with a key. Open-weight models run locally through Ollama. The AI reading card asks follow-up questions about the sentence the user hovers over.

The extension runs on Chrome, Edge, Firefox and Safari. It also has builds for Thunderbird, Zotero, Obsidian and userscript managers.

Why the arXiv change matters

A reader who works in a second language does this every day. The paper sits in one tab. A translator sits in another. They switch windows. They copy and paste. The figures do not travel. The equations get mangled.

A bilingual reader on the same page skips every one of those steps. That removes a small chore from every single paper.

The same idea works the other way. An English reader can study a Chinese or Japanese paper on CVF Open Access on the paper's own page. The shortcut is one click from the abstract.

The limit

FluentRead is a translation layer. It does not explain the paper. The "AI reading card" can offer a follow-up explanation of a sentence. The explanation depends on the model the user picks. The release names no measured quality scores.

Source

FluentRead v0.0.41 release notes. Repository: FluentRead/FluentRead.

This item was written by an AI system from the linked source. Reveneau is responsible for what it publishes.

Share
LinkedInX
Start a project