Frequently Asked Questions

Hello Zenno is an open-source web application for intermediate and advanced language learners, designed to assist with vocabulary acquisition and listening practice through AI-generated content. It focuses on helping learners expand vocabulary and improve listening comprehension by working with authentic source materials and AI-generated audio.

Intermediate-ish learners who can already parse simple sentences but need vocabulary depth & listening speed. Hello Zenno is designed for learners at roughly A2 level and above who want to expand their vocabulary and improve listening comprehension by working with authentic materials.

Hello Zenno currently supports 32 languages:


Our language coverage is driven by what Claude can handle confidently, and we're constantly working to improve language support quality across all languages. Each language has the same feature set, providing rich dictionary entries, audio generation, and contextual learning tools.

If your language isn't listed, please open an issue on GitHub or contact us directly. We prioritize adding new languages based on user demand and technical feasibility.

See All Languages →

No. Hello Zenno is a supplementary tool focused specifically on vocabulary acquisition and listening practice within the context of reading authentic source materials. It assumes you're learning grammar and other aspects of the language through other means (teachers, courses, textbooks). For a more complete solution, try Memrise.

No. Think of Zenno as a reading/listening exoskeleton—pair it with your favourite grammar source. Hello Zenno is a supplementary tool focused specifically on vocabulary acquisition and listening practice.

Hello Zenno is built on several key principles: authentic material-centered learning with real content, comprehension before production (influenced by Krashen's theories), rich contextual understanding of vocabulary, etymology as a memory aid, and progressive listening practice.

Instead of static definitions, Hello Zenno uses AI to generate rich dictionary entries dynamically as you encounter words. These include etymology (word origins, often a great memory aid!), example sentences showing context, comparisons with similar words, and difficulty indicators.

Explore Dictionary →

You can upload various types of content including text, photos of text, URLs, or audio files, and Hello Zenno will highlight words it thinks you'll struggle with because they're difficult. You can hover over these words to learn them in context, with rich dictionary entries that include etymology, mnemonics, example sentences, and comparisons with similar words.

Browse Source Materials →

Hello Zenno uses large language models to make judgments based on word difficulty, frequency, how commonly it appears in the particular source file, and how easy the word would be to guess from context. You can press 'Process this file' to find more words if you need additional help.

Hello Zenno features a flexible AI-generated search that's robust to spelling mistakes. You can search in English or in the target language, and it will present multiple results if there are multiple matches. This makes it slower but more flexible than traditional dictionaries.

Try Word Search →

We call it dynamic contextual dictation or "audio flashcards." You can scope the practice to a language, folder, or individual source file. The system finds challenging words in that content and generates example sentences for each, along with audio. It plays the audio for a sentence, and you can listen repeatedly, reveal the transcription, and then the translation. Your task is to understand what's being said, though there are no scores or feedback mechanisms yet. This focused listening practice helps bridge the gap between reading comprehension and auditory understanding.

Try Flashcards →

Browsing & basic use are free. AI‑generated content (dictionary entries, audio) currently run on Greg's credits; heavy users may be asked to plug in their own API key to keep it sustainable.

Right now everything you upload is visible to other users, so only use public‑domain or shareable source materials. Private uploads are on the roadmap.

Hello Zenno uses Anthropic Claude Sonnet 3.7, a cutting-edge LLM, for generating content. It uses Eleven Labs for audio generation. The backend is built with Python Flask, and the frontend uses SvelteKit. The application is hosted on Vercel with Supabase for the database.

There are currently no immediate plans for mobile apps or offline functionality. The web app could potentially be accessed via mobile browsers, but a dedicated mobile app is not on the immediate roadmap. Offline functionality is challenging due to the AI-generation component of the application.

The AI-generated content is generally high quality, especially for well-supported languages. However, all dictionary content is AI-generated, so there is a small risk of inaccuracies. Users are encouraged to report any issues they find through GitHub issues.

The best way to start is to select the language you're learning and check if there are already source materials at your desired difficulty level. Alternatively, you can add your own content by pasting a URL, uploading images, audio files, or typing in text directly. After processing, you can start reading with hover-enabled dictionary support or try the audio flashcards feature.

Get Started Now →

First, select the language you're learning. Then, either browse existing source materials or add your own by creating a source folder and adding files to it. Hello Zenno will extract text (if necessary), translate it, and highlight potentially difficult words. You can hover over these words to see their meanings, and click for full dictionary entries. You can also practice with audio flashcards generated from the difficult words in your materials.

The best way to report bugs or provide feedback is by opening an issue on GitHub at github.com/spideryarn/hellozenno/issues. You can also check the changelog to see recent updates.

No, Hello Zenno doesn't have progress tracking features. You can ignore words you've learned so they don't appear in your flashcards any more.

Centaur-sourcing is where humans and AI collaborate at scale to create things that would be very hard for either to create alone. Think of it by analogy with crowdsourcing, where a large groups of people have built the astonishing modern-day intellectual cathedral that is the Wikipedia. How could AI help with such efforts? In Hello Zenno, one potential idea would be for users to pool their API keys and resources to AI-generate rich dictionary entries and learning materials for languages they're interested in, making them available to the wider community. And then, of course, we'd need people to play the critical role of checking, confirming, improving and generally playing an editorial role.

The main way to participate currently would be by contributing your API key to help cover the costs of AI generation for languages you're interested in. In the future, there may be more ways for language experts to review and improve AI-generated content.

There are plans to potentially make the read-only part of the API freely available for developers to build upon. This would allow other applications to leverage the rich dictionary entries and audio content generated through centaur-sourcing.

Besides AI costs, the current challenges include complexity of the interface, generation speed (especially for full metadata for words), and content sourcing considerations. The project is looking for collaborators to help address these challenges.

Future plans include generating sentences that use multiple words from a source file in novel combinations, private uploading options for copyrighted materials, and enhanced community features to support the centaur-sourcing concept.

Hello Zenno is not designed for complete beginners. It works best for learners who already have basic grammar knowledge and vocabulary (roughly A2 level or above) and want to expand their skills using authentic materials.

No, Hello Zenno doesn't pretend to be a complete language learning solution. It focuses specifically on vocabulary acquisition in context and listening practice. You'll still need other resources for grammar, speaking practice, and beginner-level instruction.

No, Hello Zenno requires an internet connection as it relies on AI generation for many of its features. There are currently no plans for offline functionality.

No, currently all uploaded content is public and visible to all users. Only upload public domain or freely shareable source materials. There are plans to add private upload options in the future.