Readr: A Free, Open-Source AI E-Book Reader with Chat Q&A and Local Text-to-Speech

When Your E-Book Starts "Answering Questions"

Have you ever wondered, while reading a hard sci-fi novel, how much of the science in the story is actually real? When reading Project Hail Mary, could the "Astrophage" the protagonist describes actually exist? In the past, you'd have to pause, open a search engine, and leave the story world entirely to find an answer.

Readr wants to change all that. This AI e-book reader, launched on Product Hunt, bills itself as "an e-book reader you can have a conversation with." Highlight a line of text, ask a question, and the answer appears right in the page margin — with a clear distinction between what comes from the book itself and what draws on real-world scientific knowledge.

This design elegantly addresses a long-standing friction in the reading experience: the tension between curiosity and immersion. Readr embeds large language model capabilities directly into the reading interface, turning "asking a question" into a natural extension of the reading flow rather than an interruption.

Readr's Three Core Features: Q&A, Recap, and Read-Aloud

Intelligent Q&A with Full-Book Context

Readr's most impressive feature is that its Q&A capability doesn't answer in isolation — it "pulls information from the entire book." When you have a question about a plot point or concept, the AI answers with full context in mind. What's especially commendable is its honesty: when an answer draws on real-world science rather than fictional worldbuilding, it says so explicitly, preventing fact and fiction from blurring together.

For knowledge-dense genres like sci-fi and popular science, this feature is particularly valuable. It's like having a dedicated AI reading companion who has read the whole book and is always ready to help.

Delivering "full-book context" Q&A is technically non-trivial. Large language models have a limited context window, and a full-length novel often exceeds it. The standard solution is Retrieval-Augmented Generation (RAG): splitting the book into chunks, building a local vector index, retrieving the most relevant passages when a question is asked, and feeding those passages along with the question into the model. Readr's emphasis on "pulling from the entire book" strongly suggests it runs on this architecture locally on-device. This also explains how it can distinguish between "fictional worldbuilding" and "real scientific knowledge" — when retrieval results come solely from the book's text, the model can more reliably identify the boundaries of its information source.

Spoiler-Free Progress Recap

"Losing track of the plot" is a common frustration with long-form reading, especially when you pick up a complex novel with a large cast after a break. Readr offers a "recap" feature that lets you ask for a summary of "what you've read so far," with an explicit promise of no spoilers — it only covers what you've already read, never revealing what comes next.

This detail reflects genuine respect for the reading experience. Spoilers are the most sensitive landmine for readers, and the ability to technically constrain the recap to already-read content shows careful context management.

Natural Text-to-Speech, Fully On-Device

Tap the "Listen" button and Readr reads aloud from the current page in a natural-sounding voice. The key distinction, as the developers emphasize, is "on your device, no cap" — playback runs entirely on your local device with no usage limits. This stands in sharp contrast to cloud-based TTS services that charge per character or minute, making local read-aloud a genuinely zero-cost everyday feature.

On-device TTS is worth highlighting because natural, fluent speech synthesis has long depended on cloud services. Traditional local TTS sounded robotic and choppy, while cloud solutions were subject to network quality and usage limits. Since iOS 16 and macOS Ventura, Apple has significantly upgraded its system-level Enhanced Voice, enabling near-cloud-quality natural speech entirely offline after a one-time download. Readr's claim of unlimited local read-aloud almost certainly leverages Apple's AVSpeechSynthesizer or an equivalent system TTS interface, delivering high-quality narration with no additional cost.

Privacy and Open Source: Readr's Core Philosophy

At a time when most AI apps rely heavily on cloud infrastructure, Readr has taken a decidedly different path. The app explicitly states: No account, no server. Your reading data and questions never leave your device.

Furthermore, Readr is completely free and open source. Its Product Hunt listing combines the tags "Open Source," "Artificial Intelligence," "GitHub," and "Books" — a combination that clearly communicates its product philosophy: merging AI capability with open-source values and privacy protection.

For users who care about data sovereignty, this "everything local" privacy-first design is deeply appealing. You're reading files you own — EPUB and PDF — and all AI processing happens on-device. The entire pipeline requires no personal data to be handed to any third party. In the e-book space, this is a relatively rare but encouraging approach.

"No account, no server" means Readr must deploy full LLM inference capabilities on the user's local device. Two years ago, this was nearly impossible. Today, the edge inference ecosystem has matured considerably. Apple Silicon (M-series and A-series chips) features a Unified Memory Architecture that allows the GPU and CPU to efficiently share memory, dramatically lowering the barrier to running small quantized models (typically 1B–7B parameters) on-device. Local inference frameworks like llama.cpp and MLX have further simplified integration for developers. Readr's open-source nature means users can audit its no-upload claims for themselves — providing a verifiable foundation of trust rather than relying solely on a vendor's word.

Supported Platforms and File Formats

Readr currently supports iPhone, iPad, and Mac, covering the major devices in the Apple ecosystem. It opens EPUB and PDF files that users already own, with no lock-in to a specific bookstore or format ecosystem, allowing free import of any legitimately acquired e-book.

On Product Hunt, Readr received 91 upvotes, 6 comments, and ranked #9. Not a viral hit, but for a free, open-source, privacy-focused indie product, this performance suggests it has precisely met the needs of a segment of dedicated readers.

The Trend Readr Represents: Contextual AI

Readr's arrival reflects a trend worth watching: AI is gradually moving from "standalone chat tools" into specific content consumption contexts. Rather than having users switch back and forth between a reading app and ChatGPT, the smarter approach is to embed AI natively into the reading experience itself. This "contextual AI" thinking may well become a standard feature of many vertical applications going forward.

At the same time, Readr's commitment to local-first, open-source development offers a counterpoint worth reflecting on: amid the large model arms race, not every AI application needs massive cloud computing power. For relatively lightweight tasks like reading assistance, local inference paired with thoughtful interaction design can deliver an excellent experience — while also solving privacy and cost problems in the process.

Conclusion

Readr isn't a product chasing feature bloat. It focuses on going deep on one specific need: the curiosity that arises during reading. Full-book context Q&A, spoiler-free recaps, local text-to-speech — all three features serve the same goal: making reading more continuous, more enriching, and less interrupted.

Coupled with its free, open-source, privacy-first positioning, Readr offers a genuinely thoughtful choice for readers who want AI to enhance their reading experience without sacrificing data sovereignty. Whether you're an Apple ecosystem user or a hard sci-fi enthusiast, this AI e-book reader is well worth trying.