Epubtranslator Review: strong product, clear fixes (75/100)

Fact	Detail
Domain	epubtranslator.net
Category	Complete-book EPUB translation and browser reading
Pricing	Usage-based; one-time Plus packs from 8.99 to 29.99 USD
Platform	Web
Pages crawled	40
Crawl date	2026-10-16

Epub Translator makes a clear promise: translate a complete EPUB in the browser while keeping chapters, images, formatting, and reading order together. The public score table supports that position, with strong scores for first impressions and positioning (81.8/100), audience and messaging (81.7/100), usability (88.2/100), accessibility (86/100), performance (85.7/100), and technical SEO (83/100).

The product is easiest to understand when a reader is ready to add an EPUB. The main opportunities are more practical. Design execution scores 43.6/100, decision-support surfaces score 48/100, writing quality scores 72.3/100, editorial QA scores 68.4/100, and docs and self-serve help scores 70/100. The pricing page shows the available packs, but does not clearly tell readers which pack fits their usage. The result is a capable, focused workflow with a stronger product explanation than buying guidance.

Epub Translator is positioned clearly as a browser service for translating complete EPUB books while keeping chapters, images, formatting, and reading order together. The homepage makes the workflow concrete: upload one book, translate readable chapters as one job, then download translated and bilingual EPUB editions. Its strongest distinction is structural rather than simply language coverage. The free-to-paid path is less clear: the surface shows 20 Basic credits and calls the service free, while pricing metadata describes one-time Plus packs of 20, 50, or 100 credits for $8.99, $17.99, or $29.99.

A reader, language learner, or researcher can understand the job quickly: upload a complete EPUB, choose a language, preserve the book’s structure, and read or download the result. The homepage and FAQ answer the core what, who, how, and why with concrete language, including complete-file processing and translated and bilingual downloads. The audience breadth remains coherent because every path ends in the same job: keep reading the book. Add privacy, retention/deletion, support, and failed-translation answers alongside the existing operational FAQ.

Epub Translator makes the core task legible quickly: the homepage presents “Add an EPUB,” while Pricing, Log in, and Sign up are direct destinations. The three pricing cards are easy to scan, but they do not connect 20, 50, or 100 credits to books, pages, or a small-team workload, so choosing a pack requires inference. Guidance exists in the homepage’s three-step “How to Translate an EPUB Online” section and FAQ, yet there is no Help or Guides destination or search box; visitors must scroll the long homepage. Mobile parity is good in the supplied evidence, with the same headline, upload flow, FAQ, and CTA in a readable single column. Add usage examples, anchored help links, and explicit names for upload and language controls.

The public templates have a solid semantic baseline: the homepage and EPUB reader expose header, nav, main, and footer landmarks, and the captured sign-in email control is named “Email address” with autocomplete=email. Desktop and 390px mobile captures show the homepage reflowing without visible clipping. The verified issues are actionable. PSI reports failed label, select-name, label-content-name-mismatch, color-contrast, and heading-order audits; the rendered heading sequence jumps from h2 to h6 in the sampled pages. The reported 13/17 unnamed-control figure is only a candidate count and is not treated as a defect count. Run an accessibility-tree check, give each upload and language control a programmatic name, correct exact contrast pairs, and repair heading levels without changing the visual hierarchy.

Epub Translator has a coherent cream, brown, coral, and charcoal visual language, repeated rounded cards, and no measured horizontal overflow at 390px. The execution score is pulled down by interaction sizing and shared template issues. The mobile layout facts report 31 small tap targets on the homepage and 30 on pricing; the homepage menu control is 38×24px and the interaction set includes 30px controls. Normalize non-exempt controls to at least 44×44px with 8px separation. Lighthouse also fails label, select-name, heading-order, and color-contrast audits. The CSS inventory counts 58 colors, 23 font sizes, 30 radii, and 49 spacing values, but those counts include unused rules, so use computed-style evidence before consolidating tokens. Layout stability is a strength: CLS is 0 on mobile and 0.001 on desktop.

Mobile performance is usable but misses the strongest targets on the homepage and blog template. The homepage field sample records LCP 2.625 seconds and INP 255 milliseconds, both just beyond the good thresholds, while CLS is 0.03. Lab results expose a larger critical-path problem: homepage LCP reaches 8.36 seconds and the blog article reaches 7.49 seconds. PSI identifies LCP request discovery and image delivery issues on the blog, with an estimated 133 KiB saving; the homepage shows 325 KiB of unused JavaScript and 310 ms of render-blocking savings. The homepage loads 1,699 KiB over 49 requests with 15 blocking requests. Find the LCP request first, split route JavaScript, and defer non-essential third-party work.

The strongest copy states the product, audience, file type, languages, and outputs immediately: translate complete EPUB books, keep chapters, images, and formatting, then read or download translated and bilingual editions. The reader page adds useful specifics such as 80–150% text-size settings and remembered chapter choices. Keep that concrete style. Across product pages, claims about chapters, images, reading order, browser access, and privacy recur in modular blocks. Give each page one distinct job and replace repeated assurances with a proof section. The crawl also reports three images without alt text on the manga translator page; add concise alternatives for informative images.

The pricing surface is easy to scan: 20, 50, and 100 Plus credit packs show one-time prices and per-credit rates across the supplied desktop, mobile, and Traditional Chinese views. It does not yet help a reader decide. Every card repeats universal assurances such as supported size, never-expire credits, no subscription, and translated or bilingual downloads, but none states who the pack suits, what workload it covers, or what trade-off separates it from the others. Move universal benefits above the grid, then add concise best-for guidance and one usage example per pack. Keep selection overridable. The decision layer can remain simple; an interactive comparison tool is unnecessary. The pricing destination itself also needs reliable rendering before it can carry this commercial job.

There is no evidence of a current sitewide traffic suppression. The homepage returns 200, robots.txt allows crawling, the sitemap is verified, the soft-404 probe behaves correctly, and representative public pages render content in the initial HTML. The domain was registered on 2026-08-04, so the appropriate frame is growth protection rather than recovery from a measured historical decline. The concrete risk is /pricing: it returns 200, has a self-canonical and sitemap entry, yet is marked noindex, nofollow. Resolve whether pricing should be discoverable, then align the directive and sitemap. The trailing-slash variant /pricing/ also returns 200 without redirect. Standardize the slash policy and establish a baseline for indexed pages, impressions, clicks, and pricing performance.

The sampled corpus is competent and specific: the homepage FAQ and three-step HowTo align with visible copy, while product pages provide concrete limits and actions. Treat that as a release-blocking verification task, not evidence that the claims are false. Localized blog metadata adds a mechanical issue: Spanish, Portuguese, and Korean variants use generic titles such as “Blog,” and the crawler flags duplicate blog titles. The repeated missing-alt findings across product and legal pages point to a shared layout or asset-pipeline fix. Restore pricing, write locale-specific titles, and revalidate FAQ and HowTo schema against visible content.

The homepage FAQ and llms-full.txt provide useful self-serve facts about EPUB translation, pricing, supported languages, file limits, DRM, reading, and downloads.

Technical health is generally sound: robots.txt allows crawling and declares the sitemap, the verified sitemap contains 55 URLs, the guaranteed nonexistent URL returns 404, and HTTP and www roots each make one 301 hop to the HTTPS apex. Sampled public pages render their main content in initial HTML, and localized paths expose hreflang arrays. The main search-facing defect is /pricing: it is status 200, self-canonicalized, and present in the sitemap while marked noindex, nofollow. Choose whether it should rank and align all three signals. The /pricing/ variant also returns 200 without redirect, creating duplicate exposure. Standardize the slash convention and verify HSTS at the production edge; the supplied headers show TLS and nosniff but no strict-transport-security header.

Epub Translator is a good fit for readers, language learners, and researchers who need to translate complete EPUB books without rebuilding the book structure by hand. The strongest case is concrete: the browser workflow handles the complete book, and the translated output keeps chapters, images, formatting, and reading order together.

Three fixes would make the decision easier. First, pricing needs usage guidance so readers can choose between 20, 50, and 100 Plus credits with less guesswork. Second, the weak design-execution score points to mobile presentation details that need polish. Third, the site would benefit from clearer self-serve documentation and publication-level editorial QA. At 75/100, the core proposition is credible; the next gains are in choice architecture, presentation, and support.

This is a 13-dimension review based on a crawl of 40 pages completed on 2026-10-16, alongside the supplied product facts and public dimension score table. The review covers positioning, messaging, usability, accessibility, design execution, performance, writing, decision support, risk and stability, editorial QA, self-serve help, and technical SEO.

Dimension 13, Review-content integrity, was excluded as not applicable. Google Search Console was not connected, so search-enrichment conclusions are limited. The pricing facts describe free Basic credits and one-time Plus credit packs; they do not imply a subscription. See How SiteList scores for the review method and data notes.