MiMo TTS Read-aloud v1.0.7

Features

• bundle stradichenko's kindle-gst-play for Kindle playback (with notices) (#8) (9c6dae0)

Documentation

• add preview screenshots and demo recording to the READMEs (#4) (4216fb2)
• add forum-scale screenshots for external posts (#5) (dea7073)
• Kindle support status & roadmap; probe the audio environment (#6) (48c5a1b)
• Simplified Chinese version of the Kindle roadmap (#7) (0305b1f)
• sync the READMEs with bundled Kindle playback (#9) (978b524)

Full changelog: v1.0.6...v1.0.7

MiMo TTS Read-aloud v1.0.6

Features

• separate AI speed from player speed (#3) (0b89932)

Full changelog: v1.0.5...v1.0.6

MiMo TTS Read-aloud v1.0.5

Fixes

• align OpenSL ES FFI bindings with the real NDK ABI (#2) (6b21a87)

Full changelog: v1.0.4...v1.0.5

MiMo TTS Read-aloud v1.0.4

Fixes

• generate the release tag inside the same action run (#1) (5300bd0)

Full changelog: v1.0.3...v1.0.4

MiMo TTS Read-aloud v1.0.3

Changes

• Auto-generate versions and changelogs at release time (cf7c1c6)
• Adopt a pull-request workflow (e228deb)

Full changelog: v1.0.2...v1.0.3

MiMo TTS Read-aloud v1.0.2

Changes in v1.0.2 (2026-10-30)

Added

• release GitHub Action: pushing a v* tag builds and publishes the GitHub
  release automatically (changelog section as notes, plugin zip attached)
• local-only CI gate: make ci (syntax + catalogs + test battery) installed as
  a git pre-commit hook via make install-hooks
• the full test battery now lives in tests/ (unit/integration suites, E2E
  runner and driver, mock MiMo API, mock libOpenSLES.so) with
  make test / make e2e

Fixed

• release notes extraction is self-contained so tags that predate
  tools/release-notes.py build correctly

MiMo TTS Read-aloud v1.0.0

Initial release of the MiMo TTS read-aloud plugin for KOReader.

Changes

• 2026-10-30

Initial release.

Added

• Listening: continuous listening from the current page to the end of the
  book (automatic page turns), current page only, or the text selection;
  next/previous segment navigation; dispatcher actions for gestures and keys
• Read-aloud underline: the sentence currently being spoken is underlined;
  the underline is removed as soon as it has been read and moves to the next
  sentence (EPUB/TXT/PDF/DjVu)
• MiMo TTS models: mimo-v2.5-tts (9 preset voices), mimo-v2.5-tts-voicedesign
  (voice from a text description), mimo-v2.5-tts-voiceclone (voice from an
  mp3/wav sample) — with style tags, natural-language style prompts and speed
  control, optional streaming synthesis (pcm16)
• Gapless playback: sentence-level segments with 3-sentence look-ahead
  synthesis and cross-page prefetching; content-hashed audio cache
• Playback on every device: Android native OpenSL ES backend (LuaJIT FFI
  over libOpenSLES.so, no external binaries), command line player
  auto-detection (mpv / ffplay / mpg123 / madplay / ogg123 / sox / aplay /
  paplay / afplay), custom command override, automatic mp3/wav format
  negotiation, playback diagnostics
• Top menu icon: 🔊 icon in the reader's top menu opening the plugin menu
  (registered via KOReader's reader_menu_order.lua, non-destructive)
• i18n: menus and settings follow KOReader's language; English source
  strings with a shipped zh_CN catalog and a gettext workflow
  (make pot / make po-update / make mo)
• Docs: README.md (English) and README.zh-CN.md (简体中文)
• Quality gates: unit/integration suites (chunking, API payloads, playback
  backends, text location, i18n, menu wiring) and E2E runs against a real
  KOReader process with a contract-validating mock MiMo API (command and
  opensles scenarios)

Install: download mimo-tts.koplugin-v1.0.0.zip, extract the mimo-tts.koplugin directory into your KOReader plugins directory and restart.