Changelog

V3.17 — Rebuilt Tags, Sharper Overrides & a Faster Glossary

August 14, 2026

Tags Rebuilt

• Tags are now sorted into nine groups — Genre, Relationship, Protagonist, Power System, Worldbuilding, Beings & Factions, Socio-Political, Narrative and Story Element — instead of one undifferentiated pile. The chips are legible on light backgrounds now too
• Romance and danmei tags exist at last — Boys' Love, Girls' Love, Heterosexual, Omegaverse, Sentinel and Guide, Enemies to Lovers, Slow Burn, Yandere and more. Orientation is applied only on real evidence of the central pairing, and never changes the protagonist's gender
• Also added: Quick Transmigration, Transmigration into a Book, Cannon Fodder, Time Loop, Multiple POV, a full Socio-Political set (Palace Intrigue, Kingdom Building, Politics, Royalty…), protagonist-personality tags (Cunning, Pragmatic, Hiding True Abilities, Strong Female Lead…), plus Medical, Cooking, Sports, Music, Parenting, Zombie, Dungeon and Bloodline
• Tags are chosen from the whole novel, not just the opening — a story whose kingdom-building starts at chapter 300 can be tagged for it now
• Tags that only repeated the category or fandom are gone, freeing those slots for something that describes the story
• A tag now means the story is about the thing, not that it appeared once
• Chinese source-site tags are understood (耽美, 百合, 快穿, 穿书, 宫斗, 炮灰, 女强…) instead of discarded
• The whole catalogue has been re-tagged, most-read first, and now refreshes on a slow rolling basis
• Novel synopses were rewritten to describe the actual book rather than the same interchangeable plot recap

Reading

• Pull-to-next-chapter reworked — the bar thickens and lights up once you've pulled far enough, a short buzz confirms it on phones, and releasing early slides it back. It behaves the same across browsers, and on the last chapter it no longer fills up and does nothing
• Retranslating a chapter keeps your place — the new text swaps in paragraph by paragraph while you carry on reading. Same for the free machine translation
• Rescraping keeps your position too, instead of reloading from the top
• Retranslation and rescraping no longer stall or fail on long jobs, and failures give the real reason
• A finished retranslation appears immediately instead of waiting for you to touch something
• The mobile paragraph menu no longer jerks into position after opening
• Tapping outside a glossary term no longer scrolls the chapter
• Streamed translation starts arriving sooner on shorter chapters
• The reader toolbar no longer shifts as buttons appear, and "Explore the wiki" sits beside the end-of-chapter feedback box

Glossary & Codex

• The glossary dialog opens and closes instantly — it had crawled on main characters, whose chapter-appearance lists ran to thousands of entries
• Mid-story overrides can now apply to a single chapter. Corrections always ran to the end of the novel, so a one-off — a reveal, a disguise, one bad line — meant guessing where it should stop. Pick "This chapter only" or "From here on"; markers read "Ch.412 only" vs "Ch.412+". Chapter 1 works now too
• Terms discovered partway through a novel are applied backwards. A term first found at chapter 400 never existed for chapters 1–399: some already use the right name but were never linked to it, others use a name the AI invented on the spot. The first are linked, the second queued for retranslation
• Renaming a term can now retranslate instead of find-and-replace — a blind swap can't fix grammar, so "Information Beings" → "Information Being" broke every real plural
• Wiki refreshes had silently stopped for a month and are back, most out-of-date first. Wikis were also being cut off mid-write
• Suggesting a term that already exists now offers to open it instead of a dead-end error
• Change history is always reachable, and no longer shows one term's history under another's name
• An override that changes nothing is refused with a reason instead of vanishing on save
• Personal translations work on the novel glossary page — the shortcuts were empty there and edits could be discarded
• The chapter-override editor no longer flashes black and white on the way in and out

Novel Pages

• The Male Lead / Female Lead badge is clickable, filtering browse to that kind of story like the badges beside it
• Novel pages link out to the book on DaoSearch where we could confidently match it. Matching is strict — a button that confidently sends you to the wrong book is worse than none

New Source

• Added fanqienovel.com (番茄免费小说), one of the largest Chinese platforms

Source & Metadata Fixes

• 52shuku chapters had no titles — every one read "Page 1", "Page 2". Real headings inside the text are now picked up where they exist
• Novels with an emoticon in the description no longer get stuck — a kaomoji was being read as untranslated Chinese
• A novel that prints its author inside the description no longer imports blank
• Word counts had frozen on the most-read novels. Every page view counted as a "touch", so popular novels never came up for recalculation — some sat at zero despite hundreds of translated chapters

Ratings, Reviews & Requests

• The "write a review" prompt shows the review box now — it had been rendering the rating step instead, and burning its own chance to ask you again. That has been reset
• The rating prompt highlights the rating you already gave
• Queue Status and Currently Processing show their counts on arrival, and update right after you submit
• Arriving from a DaoSearch link gets your results faster

V3.16 — Source Search, Reliable TTS, Faster Pages & Smarter Codex

August 11, 2026

Find & Request Novels More Easily

• Search by novel name directly on the Request page instead of finding and pasting a source URL yourself
• Results from different websites are grouped under the matching novel, with title and author checks helping filter out unrelated books
• DaoSearch links can hand the title and author straight to the Request page; if automatic search finds nothing, a regular web-search option remains available
• Your search allowance is shown only when it is running low, keeping the normal request flow clean
• Search allowances — Translator Base 160/month, Translator Pro 480/month, plus 5/month for free accounts to try it. Repeat searches are served from cache and don't count against your allowance
• Searching for a novel we already have now recognises it straight away, instead of running a full source search first and spending part of your allowance to be told it exists. Titles typed with full-width punctuation (：，) now match the same novel stored with regular punctuation

Five New Sources

• Added support for czbooks.net, piaotia.com, twbook.cc, ttks.tw and sto55.com, along with additional mirrors for existing sources
• Improved chapter discovery and text cleanup across min-yuan, 69shuba, xsw and biquge
• Unavailable websites are temporarily skipped instead of repeatedly delaying the entire scraping queue
• Novels wrongly marked "Completed" are now Ongoing again. biquge.tw labels most of its catalogue complete — 188 of the 347 novels we hold from it — including books that gained a chapter the same day. We no longer trust that label, so novels from it start as ongoing and genuine endings are detected from the chapters themselves. If you dropped one of these because it looked finished, it's still running
• New novels now have a better cover fallback chain using DaoSearch, the source website and official-store matches before a moderator is asked for help
• wcshuba, biquge.tw and uukanshu.cc are working again. Those sites had blocked a range of the addresses we scrape from, and their block page was being read as a normal page — so chapters looked missing instead of being retried from an address the site still accepts
• Chapters that sfacg withholds (paid or removed at the source) now say so plainly instead of reporting a confusing failure
• Source status page cleanup — sources we no longer support have been removed, and a source that simply stopped being checked no longer sits at a permanent "Down"
• Genuinely gone for good: drxsw.com, qushucheng.com, trxs.me, ffxs8.top and 116xs.com no longer exist as websites, and dxmwx.org and uuks.org are broken at source

Smarter Translation & Metadata

• Every AI task now has a pinned model and a defined fallback chain. The models themselves are unchanged from V3.15 — live reading stays on Gemini 3.1 Flash Lite for speed, background back-catalog work stays on GPT-5.6 — but each job now declares what it is (reading, glossary, wiki, catalogue) and the system picks the model and effort level for it, falling back automatically when a provider is slow or down
• Character and glossary detection now runs at a higher reasoning setting, improving name and term consistency. This applies to background processing only, so live reading speed is unaffected
• Wiki writing runs at the highest reasoning setting, and wiki quality checks are now judged by the same model that wrote them
• Novel descriptions, tags, category, fandoms and protagonist gender are generated together from better source context, while moderator overrides remain respected
• Chapter titles are translated with the chapter body for better context, and chapter titles can now be retranslated on their own
• Translation retries preserve completed work more safely, with improved paragraph formatting, plural-name matching and glossary consistency

Codex & Entity Fixes

• Wiki generation now checks for same-novel duplicates and junk entries before creating a new page, with safer cleanup for existing duplicates
• General entities can no longer be hidden across every novel by one novel's setting; exclusions are reversible and can be customized per novel
• Lowercasing or renaming a term no longer changes longer terms containing it. Hiding "Hairpin" was also rewriting "Red Gold Hairpin" — a different entity. Around 5% of a novel's terms sit inside a longer one, so this was common
• Plurals now follow. Lowercasing "cushion" also catches "Cushions", and a rename reads correctly — "cushion" → "pillow" turns "Cushions" into "Pillows"
• Lowercasing a term no longer changes chapter titles, which are capitalised by convention. A genuine rename still updates them
• Turning a term off for one novel is respected while reading. A matching general entity was quietly taking its place during chapter translation, so the term kept appearing after being hidden
• Excluding a general entity from a novel now updates chapters already translated with it, instead of only affecting future chapters
• Excluding and re-enabling a general entity now follow the same review path — previously it could be switched off instantly but needed approval to switch back on
• Requested wikis and approved wiki reports are handled ahead of routine background work, with a notification when generation finishes
• Entity artwork now uses the entity's own wiki for context instead of relying only on the novel description
• Large entity and wiki pages load faster after several database-query and caching improvements

Faster, Lighter Images

• Novel covers, entity artwork and avatars now provide several WebP sizes so each screen downloads the closest useful version
• Hero images are loaded early, while images below the fold wait until they are near the viewport
• Image dimensions are reserved before loading to reduce layout shifts, with the original image retained as a fallback
• Avatar and image handling uses less memory during long browsing sessions and large community feeds

Reader & TTS Improvements

• Reader controls were rebuilt as cleaner mobile-friendly top and bottom bars with TTS, retranslation, glossary, chapter-list and navigation actions
• Added an in-reader chapter list with search, sorting, current-chapter jump and smooth handling of very large chapter lists
• Reader messages now distinguish between login requirements, tier access, source loading, translation and rescraping more clearly
• AI narration is smoother between paragraphs through better audio preparation and caching
• Browser and AI narration now recover when the next paragraph silently fails to start, fixing playback that stopped after one paragraph or failed to auto-advance
• Browser TTS no longer skips sentences or races ahead while the tab is hidden; lock-screen and next-chapter playback are more reliable
• TTS follows personal entity-name changes, while eligible patrons default to AI voices and everyone else defaults to browser voices without overriding saved choices
• Pinch-to-zoom works again — the mobile layout had been blocking it
• New setting: community glossary suggestions (Display Settings → Glossary). The blue underline on terms other readers have translated can now be switched off; with it off, those terms stay tappable so you can still set your own translation
• The retranslate model picker is temporarily hidden. Chapter retranslation now always runs on Gemini 3.1 Flash Lite while we settle the routing above; the setting will return
• Choosing the optional free machine translation now applies for the current session only, instead of sticking and re-translating chapters that already have a saved AI translation
• Paragraph retranslation now works on chapter-heading lines. Some sources paginate by fixed-size sections rather than by chapter, so a chapter can contain a heading partway through. Retranslating one used to fail with a message telling you to run a full retranslation, which could not help either — those lines are now handled as headings

Codex, Community & Account Polish

• Entity cards were redesigned as responsive portrait cards, with deferred images and a swipeable character carousel showing more of a novel's cast
• Entity likes now update immediately across chapter and Codex views
• Review replies can now be liked, with cleaner comment/review actions and smoother reply threads on mobile
• Novel reports now group missing chapters, ordering, source quality and retranslation issues into a clearer two-step flow
• Sidebar sign-in goes directly to Patreon, while actions that require an account show a shared explanation without discarding entered text
• The Original tab is now hidden for signed-out visitors rather than shown locked
• Home page shelves reordered — the Discover block (Top Rated / Top Completed) has been retired and New Arrivals promoted to a shelf of its own, alongside Trending Fanfics, Trending OGs, Trending Characters, Recent Comments and Latest Updates
• Browse, Library and Uploads now show up to 50 per page (previously 100), so a page of results loads far fewer covers at once. A saved preference of 100 is adjusted to 50 rather than reset
• Fandom cleanup — 247 novels filed under "Primordial World" now sit under Investiture of the Gods, so the two no longer split the same fandom across separate filters
• Rating and review prompts reappear sooner after being dismissed, and no longer stay suppressed once you've removed a rating on another device

Faster First Loads & Better Reliability

• Home data and important images can begin loading before the main app starts, improving landing and novel-page first paint
• The page header appears before the full interface mounts, routes warm after load, and the Inter font is delivered as one smaller variable-font package
• Non-essential dialogs, sidebars, Markdown and below-the-fold content stay out of the initial loading path
• Old cached files after an update now fail cleanly instead of producing a blank page, allowing the app to recover automatically
• Chapter, queue, health, glossary, popular and community requests were made lighter through narrower queries and better caching
• Approved missing-chapter reports now trigger a focused repair automatically, and long-session SEO, media and browser-storage growth were tightened
• New chapters are picked up sooner. The catalogue is now checked twice a day instead of once, so a novel that updates shortly after a check no longer waits until the next day. Novels with no new chapter for a month drop to weekly checks, which pays for the extra pass

V3.15 — GPT-5.6 Translation, Codex Wiki v2, Entity Likes & the Tier Refresh

July 16, 2026

New Pricing & Tiers

• Simpler round pricing — tiers are now $5 / $12 / $25 (Reader / Translator Base / Translator Pro), and existing patrons are being moved to the updated prices too (Translator Base $9.99 → $12)
• AI Text-to-Speech now on Translator Base ($12) and up — unlimited. Previously Translator Pro only
• Wiki generation is now a Translator Pro ($25) exclusive — 60/month
• Previews and full requests mirrored — Translator Base 32 / 32, Translator Pro 96 / 96 per month
• Translator Pro retranslation bumped — 75 chapter and 750 paragraph retranslations per day
• AI Chat retired — the feature saw little use and has been removed site-wide (frontend, backend, and its quotas)

GPT-5.6 Backlog Translation

• Background back-catalog upgrade — the dedicated backlog worker (introduced in V3.14 on GPT-5.5) was upgraded to GPT-5.6, re-translating older chapters (from gemini-2.5-flash-lite / DeepSeek), prioritising untranslated content first and sweeping by novel popularity
• Resilient background processing — the backlog paces itself and backs off cleanly under load instead of erroring, so live reading is never affected
• Title merged into the first chunk — one LLM call now translates title + body together and splits the title back out, saving a separate title-translation call (and dodging the Vertex 429 title fallback)
• Larger GPT/entity chunks (150 paragraphs) for ~90% single-pass throughput

Codex Wiki v2

• Rebuilt wiki generation — a single GPT-5.6 call (was nine), with incremental refresh, a daily regen beat, and a strict 100%-translated-indexed gate so wikis are never half-baked
• Wiki reports — flag a bad or spoiler-heavy wiki from the entity page; report history is visible to users
• Per-novel entity wikis — a general character now gets its own wiki per novel, RAG-scoped to that novel's chapters

Entity Likes & Codex Polish

• Instagram-style entity likes (heart) with live counts across every entity surface, plus a "Most Liked" browse sort
• Slug URLs for entity / codex / glossary / summaries pages, with canonical-on-load and legacy redirects
• Richer entity detail — wiki/image model attribution, a fullscreen image lightbox with mobile swipe, sibling-novel wikis, a Codex button, and an Entity Info box
• Entity art on gpt-image-2 (@ medium) — genre-driven render styles, grounded/epithet prompts, cross-entity identity fixes, and 640px thumbnails; full regen of the catalogue
• Behind the scenes: entity concept classifier — every entity gets a kind + attributes and within-novel duplicate detection, with a junk filter and an adversarial verify pass, quietly cleaning up glossaries

Community & GIFs

• GIFs in comments, reviews & replies — a KLIPY-powered GIF picker
• "Recent from the community" teaser on the home page — the newest reviews and comments (GIFs included)
• Notification cards render live images/GIFs now instead of static placeholders
• Review + comment improvements and better track-record attribution for glossary edits and novel reports

Faster, Lighter Site

• Cover thumbnails — WebP thumbnails cut the home payload dramatically (it had ballooned to ~28MB); served in card/list contexts
• Landing + community/home consolidated into a single edge-cached API call; static assets moved to cdn.randomtranslator.com
• Self-hosted Inter font, real <a href> navigation site-wide (open-in-new-tab works everywhere), and CLS / render-blocking-CSS fixes for a faster first paint
• Novel page — a Characters preview is embedded in the detail API, and the chapter list moved to a lazy in-page sub-page
• Multi-tier backend performance pass — query + event-loop wins, write-path entity diffing, an untranslated partial index, and shared-Redis list caches

Reader & Reliability

• Sort a novel's chapters by source order — the chapter list can now follow the original site's ordering (front-end + API)
• Paying patrons always get AI — the AI / machine-translate toggle is hidden on paid tiers, which always read the AI translation
• Manual refreshes jump the queue — a user-triggered refresh now runs at higher priority than background sweeps
• Entity occurrence counts corrected — occurrences are now counted from the English translated name in the finished chapter (word-boundary, case-sensitive) instead of the raw original, fixing badly inflated counts on common words
• Sturdier scraping — improved reliability on protected sites and fixed scraper health stats; DeepSeek "needs-retranslate" flagging tightened
• Notification improvements and entity-history fixes on entity creation

Housekeeping

• Stats page — per-model translation/entity breakdowns, wiki/image framework charts, and an untranslated slice
• Maintenance mode — an R2-flag-driven full-screen maintenance page + mod endpoint/CLI
• TTS playback fixes — instant start (no more 5-second stall then manual resume), smaller AI chunks, and an autoplay-unlock so the first chunk plays without a tap
• New scraper: min-yuan.com (小原文学网)

V3.14 — Paragraph Retranslation, Background GPT-5.5 & a Faster Reader

July 2, 2026

Paragraph Retranslation

• Retranslate a single paragraph in place instead of the whole chapter — tap a paragraph's action widget and just that line re-runs through the AI, "sandwiched" with its neighbours for coherence
• A soft "breathe" animation marks the paragraph while it re-translates; toasts auto-dismiss
• Available to Reader tier and up, with its own daily quota; quote/asterisk handling fixed so styled lines re-translate cleanly

Retranslation Controls

• Retranslate settings — pick how a novel-wide retranslate behaves (model, scope)
• Fresh retranslations run on Gemini 3.1 Flash Lite; "old model only" refreshes route to DeepSeek for cheap bulk back-catalog
• Genre-aware flavour prompts — xianxia/wuxia keep idiomatic flavour, urban/fanfic stay plain
• Entity-detection overrides now apply during retranslation too

Background Translation (GPT-5.5)

• Back-catalog re-translation kicked off — a dedicated backlog worker began quietly re-translating older chapters (those originally done on gemini-2.5-flash-lite / DeepSeek) on GPT-5.5, prioritising untranslated content first and sweeping by novel popularity
• Runs on a separate worker/queue so it never competes with live reading; force-model path with cooldown/back-off under load
• (Upgraded to GPT-5.6 in V3.15.)

Reader Reliability & Streaming

• Streaming translation hardened — the SSE generator owns its own DB session (no more "server closed connection" resets mid-stream), failed chunks retry, and a gap is never saved as "complete"
• First paragraph streams immediately without waiting on the title-translation call
• Global heartbeats stop silent phases from tripping the idle-abort, and the connection recovers when you background→foreground the tab
• Graceful "busy" handling when the per-worker live-translation cap is hit, so heavy traffic can't choke the backend
• Smoother free machine-translation reader — Google-Translate pipeline improvements for non-patron reading past chapter 50, and a cleaner "before you continue" upgrade prompt

Community & Requests

• Review replies show the replier's reading progress (e.g. "Ch. 128") so you can gauge spoilers at a glance
• Review reminders behave — no more nagging you to review a novel you've already rated/reviewed
• Search-bar requests honour your saved preview toggle — request as a preview or full without re-checking the box each time

Rank Movement

• Up/down rank arrows (daily / weekly / monthly) on the popular board and the novel header, from a daily rank-snapshot + hourly delta task
• Hides 0/null so there are no false "NEW" badges

Performance

• Faster API under load — uvicorn workers raised to 8, anyio threadpool to 64, and the get_current_user DB lookup moved off the event loop
• Chapter views buffered in Redis and flushed in bulk every 2 min; browse listing loads only the columns it serializes; DB transactions released before long LLM calls

SEO

• Original novel names woven across the whole meta system (in descriptions, kept out of titles)
• Chapter sitemap — the first 100 translated chapters per novel, plus a chapter bot-meta card endpoint

Patron Quotas Rebalanced

• Monthly full-translation requests doubled — Translator Base 16 → 32, Translator Pro 32 → 64
• Previews rebalanced down (still well above typical use) so the monthly balance favours the more valuable action — new novels

Codex & Wiki Fixes

• Empty-wiki root cause fixed — pgvector HNSW filtered search was returning 0 rows; enabling iterative scan restores full RAG context
• Variant original names can now share a single translated name without conflict

Sources & Scraping

• New scraper: wcshuba.com; trxs / xsjz / xsw extraction and SSL fixes

V3.13 — Free Machine Translation, Free Browser TTS, Saved-Search Alerts & Protagonist Gender

June 9, 2026

Free Machine Translation (read any chapter, no login)

• Read any chapter for free — the original-content endpoint is fully public now; no account, key, or tier needed, and no 50-chapter cap
• In-browser machine translation for free/non-patron readers past chapter 50; chapters 1–50 stay AI-translated and free for all, AI past 50 stays subscription-only
• Patrons get an AI / GT toggle on every chapter, and the choice persists
• Glossary-aware — known names are pre-substituted so they stay consistent; chapters fill in block-by-block with a progress bar (the old "X/Y" counter is gone)
• Background prefetch of the next chapter for instant navigation; fixed a 50→51 mis-route
• Retranslate on a machine-translated chapter re-runs the in-browser pass, not the paid AI pipeline

Free Text-to-Speech (browser voices)

• New free narration engine using your device's built-in voices — no login, no cost, now the default; AI voices stay patron-tier
• Speech settings: engine picker (Browser / AI), device-voice dropdown with preview, and a pitch slider. Removed the sleep timer and the "(free)/(premium)" labels
• Gapless browser playback; spoken text strips quote marks (apostrophes kept) and decodes HTML entities
• Browser reads 1 paragraph per chunk (AI stays 3); the highlight tracks the current chunk
• Mobile background fix — no more racing ahead; it stops when the tab is hidden and resumes on tap
• Reliable pause/resume and auto-advance on both engines; switching engines silences the previous one

Saved-Search Alerts & Daily Digests

• "Notify me" on Browse turns your filters into a saved alert (keep as many as you like): category, tags, status, fandom, word-count, preview, Forbidden
• Library digest — once a day for new chapters in your Reading / Plan to Read bookmarks, split by status (≤20 novels each)
• Discovery digest — once a day per matching alert, listing the day's new matches with a link to a pre-filled Browse page
• Discovery cards show tags, genre, and word count only — compact (e.g. "3.8M")
• Server-side preferences — feed toggles and digest opt-ins persist to your account; the bell badge counts all unread except muted types
• Bell shows 5 recent + a single "See all"; the Notifications page manages your saved searches

Protagonist Gender (Male Lead / Female Lead)

• Derived automatically from each novel's most-frequent named character; recomputed periodically and backfilled across the catalogue
• Shown on the novel page as a blue Male Lead / pink Female Lead badge
• Filter by lead on Browse and Popular (Any / Male / Female)
• Moderator override in the novel Details dialog; an override is frozen against the auto-recompute
• New "Wrong Protagonist Gender" report — approving it sets and freezes the override

Reading & Library

• New "Paused" reading status — between Plan to Read and Completed, for novels you've stopped but plan to resume
• Reading progress survives a rescrape — positions are re-bound to the new chapters (past-the-end clamps to the latest)
• Small full novels (under 50 chapters) now announce in the preview feed instead of the main new-novels channel

Entities, Glossary & Translation Quality

• Lowercase instead of retranslating — when deactivating an entity, lowercase its name in place instead of a full retranslation; can re-translate and lowercase in one pass; works for mods and via Discord approval
• Better sentence-start capitalization — names like "the Count" / "von Trapp" capitalize correctly; semicolons no longer start a new sentence
• Prompt tuning — verbatim-glossary enforcement, strict one-paragraph-in/out (no repeats), and decorative-bracket stripping
• Tighter entity detection — rejects bare characters and title-only descriptors; gender only for clearly named people
• Streaming translation under load — the first reader translates and the rest wait on the result; keepalives prevent dropped streams; navigating away cancels the work
• Entity dialog fixes — suggesting a name no longer makes the character vanish until reload, with proper rollback on failure
• Overrides chapter picker defaults to the chapter you're reading

Moderation & Community

• Report history is now visible to any signed-in user, not just moderators

Sources & Scraping

• New scraper: hjwzw (tw.hjwzw.com)
• xsw — merges the latest-chapters block with both paginators and walks "previous" links, fixing missing chapters
• hetushu — de-scrambler rewritten; skips the author-notes volume that was bloating translations
• Ad / promo stripping in chapter text (stylized-Unicode promos, known aggregator phrases, spam domains)
• Automatic cover fallback via DaoSearch when a source has none
• More reliable status — Completed/Ongoing mapped deterministically; Hiatus only for month-old novels with no recent chapters

Performance

• Faster entity search — restructured to filter by novel first, 6–14× faster on large novels
• Lighter reader requests — current-chapter lookups skip the full chapter text
• Index cleanup — dropped redundant indexes, added targeted ones, tuned autovacuum on hot tables

Interface

• Landing page — community-stats row, billions in number formatting, color-coded ratings, mobile scroll shelves
• Reader — an "N min read" estimate, a cleaned-up AI-upgrade dialog, and Fanfic + Forbidden toggles side-by-side
• Removed the 30-minute auto-reload on returning to a backgrounded tab
• Tip toasts are Chinese-source-accurate with inline links

V3.12 — Hands-Free Reading, Pull-to-Next & Smarter Translation Routing

May 15, 2026

Hands-Free Reading

• TTS auto-advance — when narration finishes a chapter, the reader automatically slides into the next chapter and keeps reading without you touching anything. Binge a whole novel hands-free
• The player stays alive across the chapter swap (a manual navigation still closes it, as before) and resumes from the top of the new chapter
• Gapless audio — the next audio chunk is now pre-decoded into its own element, so there's no more audible hiccup between paragraph groups. The gap was worst at high playback speed (shorter chunks = more switches); that's gone
• New Auto-advance toggle in Display Settings (on by default). Replaces the old TTS pause delay and sleep timer settings, which have been removed — auto-advance covers the real use case more simply

Pull-to-Next-Chapter

• Pull past the bottom of a chapter to go to the next one — a soft progress indicator fills as you keep scrolling/dragging at the end, and releasing past the threshold navigates
• Momentum-immune: the gesture only counts if it starts while you're already pinned at the bottom. The long scroll that carried you to the end (and all its inertia) is ignored, so chapters don't skip on a fast flick
• Works with both mouse-wheel and touch drag. Top/previous overscroll is intentionally left alone so it doesn't fight the browser's pull-to-refresh / back gesture

Reader Reliability

• Navigation cancels in-flight fetches — rapid Next/Prev clicks no longer flash stale content or land you on the wrong chapter. Each chapter fetch and prefetch poll is abortable and gets cancelled the moment you move on
• An aborted fetch from fast navigation is now treated as expected, not surfaced as an error
• Background prefetch is strictly a speedup now — if it's slow, manual navigation aborts it and fetches directly instead of blocking on it
• Every new chapter reliably starts scrolled to the top

Translation Model Transparency

• Per-step model display — chapters now show which model did what: e.g. Translation: Gemini 3.1 / Glossary: DeepSeek, instead of one ambiguous label. Surfaced cleanly on the chapter view (and condensed on mobile so it never bloats the UI)
• Under the hood the per-chapter model record moved to a structured {translation, glossary} shape; legacy single-label rows were back-filled so history stays accurate

Smarter Translation Routing

• Right model for the job — brand-new translations run on Gemini 3.1 (Vertex). Chapters whose existing translation came from the older gemini-2.5-flash-lite engine refresh through DeepSeek for both translation and glossary, which is materially cheaper for bulk back-catalog refresh without quality regression
• New translation router with separate DeepSeek-first and Gemini-first paths so the routing decision is explicit per chapter, not global
• Prefetch buffer tuned for the new routing so the next chapters are ready when you reach them

Entity & Glossary Quality

• Entity length caps — very long phrases can no longer be registered as entities (capped at 12 CJK chars / 8 words / 60 chars). Kept deliberately generous so real multi-word names still pass while sentence fragments get rejected
• Tighter entity prompt — explicitly rejects generic genre jargon (cultivation/system/magic terms, bare titles, numbered concept lists) with a "means the same in any other novel of this genre → not an entity" test. Kept general across all genres and intentionally compact to keep costs down
• Junk cleanup — ~23K low-value AI-detected entities with no chapter references purged, along with orphaned chapter-entity links. User-added entities were left untouched
• Natural-English tuning on the translation prompt (semantic over pinyin for common-noun names)

Free & Anonymous Readers

• No more prefetch dead-end — free and signed-out readers landing on a chapter flagged for retranslation now reliably get the existing translation to read, instead of being bounced into a prefetch loop with nothing on screen

Comments & Replies

• Smart reply targeting — hitting Reply on a nested reply now pre-fills an @mention to that specific reply's author (it used to always ping the top comment's author). The reply still threads under the parent — it's purely about who gets notified
• Cleaner now that pings flow exclusively through the mention system

Notifications

• Timezone fix — "5 hours ago" bug squashed. Backend notification timestamps now serialize with an explicit UTC offset, and the frontend parses server dates timezone-safely, so fresh notifications read correctly for users outside UTC (e.g. IST)
• Emoji stripped from stored notification text for a cleaner, consistent look

Scraping

• quanben.io added — full scraper with chapter-URL synthesis from the preview list
• ixdzs.tw mirror support, and trxs.cc now synthesizes chapter URLs from navigation when the site hides the chapter list
• wodeshucheng5.com mirror support added alongside the existing domain

Performance

• Metadata task skip-guard — the periodic novel-metadata pass now probes the latest chapter touch and skips the heavy chapter-stats aggregate entirely when nothing changed (~91% fewer expensive aggregate calls)
• Database tuning — per-table autovacuum tuned on the hot chapters / chapter_entities tables, a partial covering index added for the missing-translation scan, and a VACUUM ANALYZE cleared a stale visibility map that was forcing tens of thousands of heap fetches per query. Slow-route count dropped from 68 to 24 over 24h
• Router-level optimization on hot API paths; missing-task scan moved to keyset pagination
• Fandom wiki generation service refactored for a cleaner pipeline

V3.11 — @Mentions, Mid-Story Overrides & Mod Power Tools

May 7, 2026

@Mentions Across The Site

• Type @username anywhere you can write a review, review reply, chapter comment, or nested reply — autocomplete pops up while you type
• Mentioned users get a rich notification with the novel/chapter context and a quoted excerpt of what was said
• Mentions are ID-anchored — once you mention someone, the link survives even if they change their public username later
• Replaces the old "reply to OP" auto-notifications with a more flexible mention-based system (you can now pull a third person into a thread, not just notify the original author)
• Up to 5 mentions per write to keep things sane; daily comment/review caps still apply
• New /users/search autocomplete endpoint backs the mention picker (auth-required so anonymous bots can't enumerate the user table)

Mid-Story Entity Overrides

• Gender mid-story changes — for gender-bender novels (or any character revealed to be a different gender later), set "feminine from chapter 50 onwards" without breaking the chapters before that
• Active/Inactive mid-story changes — for entities like a defeated monster named "Silence" or "Extinction": active early, deactivated later so the LLM stops translating "everyone fell silent" as "everyone fell Silence". Multiple flips per timeline supported
• Both flow through the existing entity edit dialog. Auto-apply for mods, Discord-reviewed for everyone else
• Only chapters whose effective state actually flips get flagged for retranslation — no over-flagging
• New dedicated Chapter Overrides Dialog in the frontend with a richer editor for both segment types

Translation Quality

• Smarter entity translation prompt — natural English over pinyin. Personal names still transliterate (萧炎 → Xiao Yan), but common-noun-as-name now goes semantic only (妖怪 → Monster, 钢骨 → Steel Bone — never Yaoguai/Gangu). Place names with meaning translate semantically (天荒大陆 → Wild Continent, not Tian Huang). Pinyin is the last resort
• Rule 3 added: never extend a name past its actual characters. If 黎 is a surname, the LLM won't return 黎敏锐 (敏锐 = adjective "keen")
• Wiki quality restored — the April model swap regressed wiki generation from ~25K-char detailed pages down to ~7K-char stubs. Reverted the section model back to Gemini 2.5 Flash with explicit length guidance — wikis are detailed again
• 483 short or truncated wikis auto-regenerating in the background — affected entities will refresh over the next few hours

Mod & Editor Tools

• Editing general entities — mods can directly edit a general entity's translation, gender, etc. (used to be only "exclude from this novel")
• Bulk rename — find-and-replace by root word across an entire novel's entities. Each renamed entity still gets its own audit row, so the change history looks identical to N single edits. Auto-applies for mods; Discord-reviewed for everyone else
• "Old model only" retranslate flag — when triggering a novel-wide retranslate, mods can now check "only chapters whose existing translation came from the previous-generation model". Lets you refresh stale translations without redoing chapters that already use the latest engine
• Cover change via novel report — users can submit a "wrong cover" report with a URL; mods can one-click apply the new cover from Discord. Mods can also change covers directly via the moderator actions panel. New cover URL is cache-busted so users see the change immediately
• Mod log embeds now show a generic "Changes" field for any mod action, not just the ones with hardcoded fields

User Uploads Page

• New /me/uploads page showing every novel you've contributed, with formatted stats (word count, views, bookmarks, readers as 12.3K / 1.2M shorthand) and last chapter timestamp
• Distinguishes Preview vs Full uploads with a clear flag

Notification Format

• Reply notifications got prettier — replies on your reviews and chapter comments now show the original text and the reply each in their own markdown blockquote, with novel and chapter context inline. Same shape as the new mention notifications, so all conversation notifications look consistent
• Vote notifications got prettier — same blockquote treatment when someone likes/downvotes your review or comment
• Shared format_blockquote helper lives in common/utils/notification_format.py so all notification creation sites stay in lockstep

Scraping

• xsw.tw rebuild — the site's ascending and descending paginators had drifted out of sync, causing chapters to silently disappear from one direction or the other. Now we fetch both and union by chapter ID. Plus a new step that scrapes the "Latest Chapter Preview" section on the main novel page — those chapters publish there before hitting the paginated index, so the refresh task no longer lags behind the source. Net effect: ~140 previously-unreachable chapters across already-scraped xsw novels are now back
• shiguangshuwu fix — re-enabled after the rebuild fixed an extraction regression (~227 lines of scraper changes)
• biquge.tw chapter duplication fixed — the scraper was unconditionally fetching _2.html for every chapter on the assumption that biquge always splits chapters into 2 pages. For single-page chapters the source returns page 1 again, so the content silently appended to itself ("chapter restarts halfway through"). Now reads the (N / M) page marker first and only fetches pages 2..M when the chapter is actually multi-page
• kks101.com added — that's 39 supported sites total
• Cloudflare CN-language detection — the scraper was missing Chinese-language Cloudflare interstitials (正在验证浏览器, 正在进行安全验证, etc.), making some CN sites silently fail at the JS challenge step. Now correctly detects them and routes through the bypass

Reliability

• 🚨 Cloudflare Tunnel rate-limit bug fixed — after switching to Cloudflare Tunnel + Traefik, X-Forwarded-For was returning the cloudflared container's internal IP, so every user in the world was sharing the same rate-limit bucket and getting instant-429'd. Switched to CF-Connecting-IP header. If you've been seeing random rate-limit errors, this was why
• Infinite retranslation loop fixed — entity detection during translation was over-eagerly flagging neighbor chapters when the LLM "rediscovered" existing entities. Combined with prefetch, this caused chapters to retranslate on every reload. Cross-chapter flagging now only fires from explicit user-driven entity creation, not from translation runs
• Retranslate / rescrape now available to Reader tier — previously locked to higher patron tiers, now opened up to Reader ($4.99) and above. Daily quota scales with tier so higher tiers still get more headroom
• Patreon proxy manager — Patreon API calls now route through a dedicated proxy pool, fixing intermittent 403s on the verification flow
• Annual tier auth fix — annual subscribers were occasionally being treated as expired; the verification flow now correctly recognizes annual billing
• Redis cold-start tolerance — fastapi no longer dies on boot if Redis (over Tailscale) takes >5s to first respond; bumped connect timeout to 15s and added a lazy-reconnect cooldown so a one-off blip doesn't cascade into 14 separate reconnect attempts
• Container ulimits raised — fastapi was hitting the default nofile=1024 under load; bumped to 65536 in compose
• Chapter-scan index — partial covering index for the periodic "find missing embeddings" task (51 MB, planner now uses Index Only Scan). Trims periodic-scan time significantly
• Missing-embedding task switched from OFFSET paging to keyset pagination on chapters.id — eliminates O(N) cost on deep pages
• Queue-info polling cache bumped from 8s to 30s (~2/3 of polls now hit the cache); inspector broadcast timeout cut from 5s to 1s
• Embedding batch size bumped for faster bulk processing
• Periodic missing-image task added — finds entities that should have AI character images but don't, schedules generation
• Discord mod log image rendering fix
• Alembic baseline reset — squashed 73 individual migration files into one baseline_initial.py. Pure ops cleanup, zero schema impact

Performance

• Entity search rewrite — /novels/{id}/entities/search was doing a SQLAlchemy multi-join + chapter aggregation across every match before pagination. Switched to the same CTE-based pattern the unfiltered list endpoint uses, which scales much better on novels with thousands of entity matches
• only_user_translated filter added/fixed on the entity search endpoint — show only entities you've personally translated within a novel

Glossary Cleanup

• Fandom dedup — merged redundant fandom variants into rollups: all major-league sports → "Sports", Godzilla franchise variants → "Godzilla", DC sub-fandoms → "DC". Down from 138 → 125 fandoms
• Tag dedup pass with the same approach
• Better fandom and tag changes propagate to the mobile UI
• Entity image generation rewritten for cleaner pipeline

Frontend

• Sidebar navigation refresh — chat history surfaced in the sidebar, faster nav between recent chats and novels
• New ChapterOverridesDialog — dedicated 685-line component for editing chapter-bounded gender + active/inactive overrides
• Inactive override editor in the entity dialog (matches the gender override editor)
• Mid-story gender override editor added to the entity dialog
• Bulk entity update / rename UI in NovelEntitiesView (mods only) — find-and-replace flow with preview + apply
• Mention system UI — RichTextarea (515-line component) with @ autocomplete dropdown showing public_username + avatar; MentionContent.vue renders @username substrings as clickable links inside reviews/comments
• New UserUploadsView for the /me/uploads page (439 lines)
• General Entities view got an editing mode for mods (was view-only)
• Cover change UI in mod novel-report flow + moderator actions panel
• "Old model only" toggle in the mod retranslate dialog
• New site tips popup with 24 tips for first-time visitors and an uploads-view tip
• Mod badge component for showing moderator status next to usernames
• Better responsive dialog primitive — replaces a bunch of ad-hoc modal handling across reviews / comments / feedback summary
• Chapter view + chapter content improvements (live editing, layout polish)
• TTS overlay updates — default voice fix, layout cleanups
• Rating reminder buttons more responsive, dismissable toast, "don't show again" option
• Cloudflare-friendly auth flow fix — apiClient + auth.ts cleanup; cookie handling now works correctly under CF Tunnel
• API base URL migration — App.vue now handles the API endpoint migration cleanly
• Theme handling + entity highlighting fixes
• Migration mode flag turned off (we're past the embeddings migration)
• FandomCombobox simplified + better mobile experience
• Username suggestion gap fix during signup
• Edge-case fixes in entity dialog interactions, focus handling, dropdown closures
• App-level toast improvements + warning suppression cleanup

V3.10 — Faster Translations, Smarter AI, Monthly Limits

April 15, 2026

Translation Pipeline

• Upgraded the whole translation stack to Gemini 3.1 Flash Lite Preview — the same model now powers translation, entity detection, chat, and wiki generation
• New parallel chunk pipeline — the first chunk streams live while the rest translate in parallel in the background. Big chapters finish in seconds instead of minutes
• Automatic failover across multiple translation accounts when one gets rate-limited — zero visible slowdown on busy days
• New fallback chain for content-policy blocks: if the primary model refuses, we cascade through alternates and finally to Grok. Chapters that used to show "content omitted" placeholders now come back fully translated
• Fast-fail on blocked content — we no longer waste retries on prompts the model has already refused; the next model kicks in immediately
• Smoother streaming under load — the translator releases its database slot before the long LLM call, so heavy traffic no longer chokes the backend
• Tuned chunk sizes and post-processing for cleaner output

Entity Detection & Glossary

• Cleaner detection pipeline — entity mappings save as a baseline before the LLM pass, so even if detection hiccups mid-chapter, your chapter still has clickable entities
• Junk filters added — common Chinese/Japanese/Korean words and English filler words no longer get flagged as character names
• Massive glossary cleanup: removed roughly a million bloated entities that had built up from earlier over-detection runs. Glossary pages load faster and search is noticeably more relevant
• Canon/fanfic auto-promotion — when a mod marks an entity edit with reasons like "canon name" or "fandom", the entity is automatically promoted to the global Fan-Fiction glossary. 91 historical entries were backfilled based on past change history
• Disabling a general entity for a novel now correctly lets you add a novel-specific override with the same name
• Entities with middle-dot characters (like 贞特·崔格) now match properly in detection

Codex (AI Features)

• Wiki pages, character chat, and AI image generation are back online after the embeddings migration
• Wiki and character chat now use the upgraded translation model for better quality
• Dual-path chat streaming — normal chats use our fast primary path, NSFW chats route through the alternate
• Character image generation switched to a faster/cheaper tier and images are now hosted on our own CDN
• Wiki generation waits until a novel is at least 60% indexed before it can run — prevents half-baked wikis
• Character wiki pages load dramatically faster — the main query that powered them was the biggest single source of database load and got completely rewritten (also fixed a subtle bug where character stats were bleeding across novels)

Scraping

• New scraper: trxs.org — that's 38 supported source sites total
• Universal URL normalization — paste any URL from a supported site (novel page, chapter page, catalog, pagination, search) and the scraper now auto-resolves to the novel page. No more "invalid URL" errors because you pasted from deep in a reader
• Hiatus detection fix — 92 novels that were wrongly marked Hiatus (even though they were getting new chapters) are now correctly marked Ongoing
• Cleaned up a couple of novels stuck in a "0 chapters" zombie state from failed initial scrapes
• Faster scraping — a new database index cut the scraper's "does this chapter already exist?" check from ~100ms down to under 1ms, which was by itself eating 64% of our database time
• Inkstone uploader script improvements

Database & Site Performance

• Added the new chapter-URL index mentioned above — biggest single perf win of the release
• "Latest updates" query on the homepage rewritten — 100× faster (went from ~1 second per call to ~11ms)
• Character wiki entity query rewritten — another huge win in both speed and correctness
• Novel search tuned — dropped a heavy similarity scan on full descriptions that was eating 19% of database time with almost no search-quality benefit
• General FastAPI tuning across the board: fewer redundant health checks, slimmer worker pool, cleaner connection handling

Rate Limits — Weekly → Monthly Rollover

• Weekly request limits are now monthly. Scraping, previews, chat messages, and wiki generations now use a single monthly pool instead of expiring every Monday — unused quota carries over within the calendar month
• Monthly cap = previous weekly cap × 4 (Translator Pro: 32 novel requests, 128 previews, 1,000 chats, 40 wikis per month)
• Retranslations stay at 50/day — that one doesn't roll over
• Pools reset on the 1st of each calendar month at 00:00 UTC
• Existing weekly usage was migrated — your April activity counts against the new monthly bucket
• Mid-month tier upgrades apply immediately against the new higher cap

Reliability & Redis

• Fixed a bug where the /me endpoint would sometimes show your level as blank and other times show the real number — now consistent on every page load
• Your level badge updates as soon as the nightly XP recalculation runs, instead of staying stale for up to 5 minutes
• Fixed a rare race condition in the daily rate limiters for comments, reviews, and reports
• Redis cache layer now has a connection pool and faster serialization — dashboard pages load quicker
• The /me/limits page cut its Redis round-trips by 80% per load
• Slow-request tracking redesigned — the admin dashboard can now actually tell you which routes are slow, not just which unique URLs were slow once
• Fixed a timezone bug where scraper and API services could disagree on when "this week" started, very rarely letting users double up on their quota around midnight UTC

Infrastructure

• Sitemap + RSS feed generation moved from GitHub Actions (which was getting blocked by Cloudflare) to a Celery Beat task that runs every 6 hours
• The background "check for missing chapter embeddings" task now actually runs — it had been silently stalling for a while
• Grok added as the final fallback for content-policy blocks
• Discord mod bot rate-limit fixes and tag handling improvements

Frontend

• Level badges on profile pages and next to usernames across reviews, comments, community rankings, and activity feeds
• New entity list sheet with plural entity handling (e.g. "Goblins" and "Goblin" are merged in display)
• Custom fandom input — type your own fandom name when requesting a translation
• Entity dialog in chapters opens even when the clickable-glossary toggle is off
• Prefilled reasons in novel report and entity change forms
• Clickable glossary hover states and styling polish
• User engagement history view — new activity feed
• Request view cleanup — disclaimer removed, prompts prefilled
• Streaming progress visual — live progress bar while a chapter translates
• Rating reminder toast after you've read enough chapters of a novel
• Notification UI refresh
• Chapter summary fixes, text capitalization and pluralization cleanups
• Review / comment collapse in long threads
• SEO optimization + repo size trim — smaller frontend bundle, faster first paint
• Scrollable novel tags on desktop (not just mobile)
• "Category" label renamed to "Genre" site-wide
• Skeleton loading fixes across views
• Back-navigation fix in the browse view
• Auth flow / cookie handling fixes
• Optimistic update improvements across interactions

V3.9 — Levels, On-Demand Translation & Quality

April 8, 2026

User Levels & XP

• New leveling system — earn XP from reading, bookmarking, rating, reviewing, editing entities, uploading novels, and more
• Levels displayed next to usernames in reviews, comments, community rankings, and activity feeds
• XP breakdown on your profile showing contributions across all categories
• Fully retroactive — your existing activity already counts
• Community leaderboard with top contributors ranking

On-Demand Translation

• Chapters translate in real-time as you read — no more waiting for batch translation
• First 50 chapters translate automatically when a novel is added
• Next 2 chapters prefetch in the background while you read for seamless experience
• Streaming translation shows text appearing live as it translates
• Chapter titles translated in batch for full browsable chapter lists

Translation Model Upgrade

• Upgraded to Gemini 3.1 Flash Lite Preview (1M context, 65K output)
• Single-shot entity detection — full chapter analyzed at once instead of chunks
• Larger translation chunks (50 paragraphs) for better context and fewer seams
• Improved capitalization handling — proper nouns capitalized, common words stay lowercase
• Fixed streaming token spacing issue — no more mashed-together words during live translation

Chapter & Novel Improvements

• Chapter counts now show total chapters (not just translated ones)
• Word counts include estimates for untranslated chapters based on original content length
• Chapter detail API now shows when chapters were last updated
• Duplicate chapter prevention with Redis locking during refresh
• Fixed novel refresh going through wrong queue (now properly urgent for user-triggered)

Entity & Glossary

• Entity name changes from the edit dialog now go through mod approval (instead of silently saving as personal translation)
• Added user_id tracking to entity change history and novel reports for accurate attribution
• Improved fandom classification — fan-fiction novels set in existing universes correctly categorized
• 143 fan-fiction novels retroactively assigned fandoms via LLM classification

Queue & Task Management

• All task timeouts capped at 5 minutes (down from 30-60 min) — no more stuck workers
• Translation tasks split into urgent (reader prefetch) and normal (batch/mod retranslate) queues
• Mod retranslate only processes already-translated chapters (not all chapters)
• Metadata retranslation moved off urgent queue to prevent blocking user actions
• Refresh digest notifications moved to metadata worker
• Concurrent novel refresh prevention with Redis locks

Discord Bot

• Error recovery for stuck "PROCESSING" states — reports can now be retried after failures
• Novel report approval/rejection tracks user_id for accurate XP attribution

Bug Fixes

• Fixed chapters with NULL views/word_count causing 500 errors
• Fixed review timestamps showing wrong time ("23 hours ago" on fresh reviews)
• Fixed non-authenticated users triggering expensive on-demand translations
• On-demand translation now requires $4.99 patron tier (LLM cost protection)
• 69shuba.tw scraper reliability improved with a faster direct-fetch path

---

V3.8 — Images, Glossary & Performance

April 5, 2026

Original Novel Covers

• All 3900+ novels now use real cover images from source sites instead of AI-generated art
• Images served from our own CDN for faster loading
• NSFW novel covers automatically blurred in browse and popular pages

Personal Glossary (User Translations)

• Set your own translation for any character/entity name — takes effect instantly
• Your translations appear in chapter text with a dashed underline
• See what other readers chose — popular translations shown as quick picks
• If 3+ users pick the same translation, it gets proposed to mods as the new default
• Manage all your translations from the novel glossary page with "My Translations" filter

Reading Status

• Track your reading progress: Reading, Plan to Read, Completed, Dropped
• Filter your library by status
• Colored status badges in your library

UI & Design

• Custom themes support
• Contextual tips across all views and dialogs
• Improved skeletons and loading states
• Better reviews and comments design
• Mobile image size optimizations
• TTS dialog improvements

Performance & Reliability

• Cloudflare cookie caching — pages from protected sites load 10-15x faster
• Task queue deduplication redesigned — no more queue explosions
• Chapter content pre-processed at translation time — faster page loads
• HTML angle brackets in content no longer disappear
• Smart digest notifications — batched chapter update alerts instead of spam

Translation Quality

• Gemini 2.5 Flash Lite for all LLM calls (faster, cheaper)
• Canonical tag system with 110 curated tags
• Traditional/Simplified Chinese author matching — same author always grouped correctly

New Sources

• hetushu.com, 52shuku.net

Infrastructure

• All images migrated from Cloudflare Images to Cloudflare R2
• AI image generation removed — real covers only
• Cover images automatically downloaded and stored for new novels

---

V3.7 — The Community & Quality Update

March 31, 2026

Translation Pipeline Overhaul

• Entity pre-injection reworked — entities annotated as context hints, AI adapts capitalization/grammar naturally
• Per-chunk entity detection — fixes JSON errors for novels with 1000+ entities
• Stricter entity detection — no more generic nouns polluting glossaries
• Gender detection improved — characters properly classified as M/F
• CJK validation — entities with Chinese translations auto-rejected
• AI glossary entries self-correct on retranslation (user edits protected)
• Smaller chunks (20 paragraphs) for better quality on long chapters
• Lower AI temperature to reduce hallucination
• Higher output token limit to prevent content truncation
• Ellipsis normalization to prevent abrupt endings
• 14,000+ entity entries cleaned, 600+ swapped names fixed

Chapter Comments

• Full threaded comment system on every chapter
• Reply, upvote/downvote, edit, report

Review Upgrades

• Replies, upvotes/downvotes on reviews
• Vote counts and reply counts visible
• Notifications when someone interacts with your content

Translation Feedback

• Thumbs up/down on chapter translations
• Issue tags: missing content, wrong pronouns, names wrong, added content, unreadable
• Mod dashboard for worst-rated chapters

Novel Reporting

• Report wrong genre, status, title, description, author, fandom
• Discord-based mod approval workflow
• Notifications when reports are resolved

Notifications Revamp

• Vote, reply, and report resolution notifications
• Rich context with links and previews
• Removed 4.5M bloat notification rows

Glossary Multi-Category

• General entities support multiple categories
• Fandom field supports multiple fandoms per novel
• Fandom data cleaned and deduplicated

New Pricing

• $4.99/mo Reader tier added
• $9.99/mo and $24.99/mo tiers updated

New Sources

• biquge.tw, book.qq.com, xiaoshuo.com.tw

---

V3.6 — The Polish Update

March 27, 2026

Translation & AI

• Translation algorithm updated with improved pipeline
• Structured schema output for responses
• Entity detection limit removed
• Gemini model integration
• "Find original" feature for source text lookup

Frontend

• i18n/multi-language routing removed — English-only, faster
• Chapter progress tracking
• Diff chapter view for comparing versions
• Prompt page added
• Various layout refinements

Infrastructure

• Code cleanup across fastapi
• Stats API cache improvements
• New proxy rotation for scrapers

New Sources

• Book.qq domain added
• Various scraper stability fixes

---

V3.5 — The Foundation Update

February 17, 2026

Translation & AI

• Switched to Gemini Flash Lite
• New entity detection prompts
• Entity update triggers auto-retranslation
• Low-usage entity cleanup

Frontend

• Novel slug URLs — cleaner URLs
• Estimated word count display
• New themes and reader customization
• Word count filter in search
• Mark all as read for notifications

Infrastructure

• UUID to Integer database migration (all services)
• Domain migration to randomtranslator.com
• RunPod integration for GPU workloads

Moderation

• Regenerate Covers mod action
• NSFW image detection
• Discord webhook alerts for duplicates

New Sources

• Twkan, 69shuba.tw, Wodeshucheng, UUKS, 44XW, UPkj, and more

---

V3.4 — Infrastructure & Backend

December 15, 2025

• Major database and server migrations
• API speed optimizations
• Novel filter logic & caching optimizations
• Entity change history tracking
• TTS speed adjustments and bug fixes
• Codex model updates and chat improvements
• General entity exclusion system
• Site hosted on dedicated server
• Documentation site launched

---

V3.3 — TTS, Prefetch & Codex Improvements

September 2, 2025

TTS Upgrades

• Click any paragraph to start reading from there
• Fixed widget overlap between mini-chapter actions & TTS widget

Chapter & Reading Flow

• Prefetching next chapter for seamless, faster navigation
• In chapter view: Original tab shows Rescrape, Translated tab shows Retranslate
• Character previews (entities with wikis) on novel pages — generate new wikis on demand

Mod Tools

• Reset Glossary fix — now works in Retranslate & Rescrape
• Active Discord users get Mod Roles — update details, schedule retranslates/rescrapes

Codex Wiki Updates

• Regenerate Wiki button — refresh bugged/incomplete wikis instantly

UI & Site

• New homepage with trending fanfics, latest updates & arrivals
• Display Settings overhaul — cleaner, more concise look
• Fixed white background flashes when overscrolling in dark mode

Performance & Stability

• Site optimizations + caching updates — faster & lower latency
• Expanded scraping sources
• General bug fixes + UI polish

Algorithm Updates

• Search algorithm — better typo tolerance & more relevant results
• Translation algorithm — improved entity detection in long chapters for consistency
• Glossary update — Local (book-specific) vs General (site-wide), disable terms that don't fit context

New Sources

• renrenreshu.com, shiguangshuwu.com, mokakanshu.vip, ixdzs8.com, fsuzw.com, qushucheng.com, 116xs.com

---

V3.2 — Translation Previews

July 27, 2025

Translation Previews

• Request translations as Previews — translates up to 40 chapters as a trial
• 16 Preview requests/week for $10 tier, 32 for $25 tier
• Preview novels marked with a P icon on pages and lists
• Convert previews to full translations from the novel page

Other Changes

• Notification system added
• Community Ranking page
• My Limits page — view your current quotas
• My Stats page — check your activity stats
• Site layout optimizations
• Browser compatibility enhancements
• New URL source additions
• Model updates for Translation and Codex
• Short Translation Framework optimizations
• Mod feature flow changes
• Miscellaneous bug fixes

---

V3.1 — Codex Initial Release

July 8, 2025

Codex

• AI-powered Fandom Wiki style knowledge base for novel entities
• Wiki generation with character summaries, relationships, and chapter citations
• Image generation for character illustrations
• Character chat/roleplay (SFW and NSFW)
• Auto-generated wikis for top 200 translations

Codex Rate Limits (Per Week)

• Free users: 10 Chats
• $10 users: 1 Wiki Gen, 20 Chats
• $25 users: 5 Wiki Gen, 50 Chats
• 1 NSFW chat counts as 2 normal chats

Other Changes

• Home page improvements
• Chapter page accessibility additions — keyboard shortcuts for Retranslate, Glossary view, and TTS
• Fanfic popular list penalties for consecutive fandoms to increase diversity
• Translate page speed improvements and UI enhancements
• Caching enhancements for reduced latency
• New URL source additions
• Translation framework instruction changes
• Miscellaneous bug fixes

---

V3 — Official Launch

June 12, 2025

Feature Additions

• Official landing page launched at home.randomtranslator.com
• Entity Additions — click glossary highlight button to add new entities from chapter text
• Tag Inclusion/Exclusion in Browse Page
• Fandom classification — auto-detected by DiceAI based on novel context
• Translate page — free LLM-powered text translation using our framework, supports all major languages

General Changes

• Model updates to the latest offerings
• Entity translation fixes with more balanced transliteration
• Translation post-processing enhancements for better formatting
• Privacy policy and Terms of Service updates
• New mod tools
• UI enhancements
• Bug fixes

---

V2.9 — Multi-language Phase 3

June 2, 2025

Translation Updates

• Korean, Japanese, German, Russian, Spanish, Portuguese, English and French added as original supported languages
• CJK-specific fix for untranslated characters remaining in translation
• Entity filtering fixes — no longer uses common or irrelevant terms
• Fixed existing entities with untranslated characters
• Fixed paragraph repeating error

Site Changes

• Website Guide added at the top with sidebar changes
• Bookmarks and user elements shifted to profile click menu
• Site optimizations for lighter load on mobile
• Discord Bot fixes for entity updates
• Security fixes due to recent DDoS
• New showcase page for sample translations in all supported languages
• Added uploader info in ranking and browse tables
• Metadata and text changes for ongoing rebranding
• Preview logo changes
• UI enhancements and bug fixes

---

V2.8 — Multi-language Phase 2

May 24, 2025

• Novel translations now available in Russian, German, Portuguese, French, and Spanish
• Language translations count the same as novel requests (same rate limits)
• DDoS recovery fixes

---

V2.7 — Multi-language Phase 1 & Codex Beta

May 23, 2025

Codex Beta Phase 1

• Codex Beta Version 1.0 released with wiki and image generation
• Content indexing finished, version 1.5 under development

Multi-language Support

• Site interface translations in French, Spanish, Portuguese, German and Russian

UI/UX Improvements

• Sidebar genre addition
• Removed unnecessary toasts
• Responsive dialog boxes — redesigned for desktop and mobile
• New browse filters (New novels, search by Uploader)
• Activity and Community view time localization fix
• Desktop layout padding improvements
• Ranking colors addition
• Tables open items in new tabs

Performance

• Decreased server processing times
• UI consistency fixes
• Translation and scraping bug fixes

---

V2.5 — Major Site Overhaul

April 30, 2025

Translation & AI

• Upgraded to gemini-2.5-flash-preview model
• Minor translation architecture improvements
• Enhanced image model: gpt-image-1
• AI-generated synopses for cryptic author descriptions

UI/UX & Performance

• UI and UX enhancements across the site
• Caching and speed improvements for browsing
• Rate limiting for novel requests and retranslations

Community & Discovery

• Community Stats and Activity pages
• DiceAI novel recommendation integration on novel pages
• AI-powered tag and category standardization
• Better Browse page with improved filtering
• Reworked popular page algorithm for more dynamic listings
• Added uploader information for novels

Other

• Mod feature additions
• Scraper fixes
• Bug fixes

---

V2.4 — System TTS & Bug Fixes

March 31, 2025

• Added Basic System TTS accessible in the $10 tier
• Bug fixes

---

V2.3 — DiceAI Novel Search & Recommendations

March 28, 2025

• AI-powered search with knowledge of all novels in the database
• Personalized recommendations based on reading data, ratings, bookmarks, and reviews
• Full privacy control — activate/deactivate personal data usage anytime
• Premium feature for $25+ tier users
• Bug fixes and UI enhancements

---

V2.2 — AI Text-to-Speech

March 20, 2025

• AI Text-to-Speech with 27 voices and adjustable speed
• Premium feature for $25+ tier users
• Fixed refresh chapters

---

V2.1 — Translation Improvements

March 15, 2025

Entity and Translation Architecture

• Better context and consistency measures
• Entity detection now focuses on significant entities over common nouns
• Better grammatical flow of entities in translations

Quality & Performance

• Fixed irregular formatting issues for system texts
• Major speed improvements
• Upgraded to better translation model

---

V2 — Beta Launch

March 10, 2025

• Full site launch at randomtranslator.com (migrated from geriatrictranslations.com)
• Novel request system — add novels by pasting source links
• Glossary consistency with gendered entities and in-chapter edits
• Reader-friendly UI compatible with Safari Reader Mode
• Patreon login integration