Changelog

v1.14.7 (July 8, 2026)

What's new for you

• AI Assistant is now free to use — Sign in with Google and use the in-meeting AI Assistant without your own API key, up to your included daily and monthly quota. Responses now stream in real time for a noticeably smoother experience.
• Markdown-formatted assistant responses — Assistant messages now render with proper formatting — headings, lists, bold text, and code blocks — instead of plain text.
• Meeting Summary & Task Breakdown for everyone — These AI-powered exports are no longer limited to users with their own OpenAI key. Any signed-in user can generate them using their included AI quota.
• Custom OpenAI-compatible endpoint — Power users can now point translation and AI features at a custom OpenAI-compatible base URL and model from Settings.
• Multi-select transcript management — Select multiple transcript entries at once to download or delete them together.
• Included usage quota shown on the translator page — See how much of your daily/monthly quota remains without opening Settings.
• Synced translation and narration language — Narration language now automatically follows your translation language selection.
• Fixed Narrate their voice on YouTube — Resolved an issue where this feature stopped working correctly after switching translated languages.
• Fixed Google TTS language errors — Corrected unknown-language-code failures for certain Google TTS voices.
• Fixed duplicate context-menu error — Reloading the extension no longer triggers a duplicate context-menu-id error.
• Fixed duplicate page content on webpage translation — Resolved an issue where translated pages could show duplicated content.
• Facebook and WhatsApp permission fixes — Resolved permission errors that could prevent the extension from working correctly on these sites.

v1.5.3 (June 16, 2026)

What's new for you

• Optimized sign-in flow — Google sign-in is now integrated directly in the translator panel, so you can authenticate without leaving your meeting or video page. Fewer steps, less friction before your first translation.
• Prominent Start button — A clear Start action on the empty translator screen makes it easier to begin your first session. New users can see exactly what to do instead of hunting for controls.
• WhatsApp Web screen-sharing fix — Fixed an issue where the browser’s screen-sharing popup could fail or behave incorrectly on WhatsApp Web when starting translation or related audio workflows.

v1.0.6 (June 3, 2026)

What's new for you

• Translate any webpage — Right-click on a normal website and choose Live Translator → Translate this page. The extension opens a full-screen dual-panel reader: original article on the left, streaming translation on the right, with linked scroll between panels.
• Cross-panel sentence hover — Hover a translated sentence to highlight the matching segment in both panels and see the original wording in a tooltip.
• Readability-based extraction — Article content is extracted with Mozilla Readability so you read the main story without mutating the host page. Choose Exit translation to close the reader and return to the unchanged site.
• OpenAI key for webpage translate — Webpage translation uses your OpenAI API key from Settings (same requirement as AI Assistant). Soniox is not used for this feature.
• Included Soniox access — no key required — Sign in with Google under Settings to start live meeting and video translation without entering a Soniox API key. Usage counts against your Free or Pro included quotas.
• Included access quotas — Free: 20 minutes per day, 400 minutes per month, up to 4 sessions per day. Pro: 120 minutes per day, 1440 minutes per month, up to 24 sessions per day. View remaining time under Settings → Soniox.
• Bring your own Soniox key (optional) — Add your Soniox API key in Settings for unlimited translation; included quotas no longer apply while your key is configured.
• Clearer translator onboarding — When included access is available, the translator panel prompts you to sign in with Google before Start is enabled.



v0.30.8 (May 22, 2026)

What's new for you

• Auto-save transcript site allow list — Settings → Transcript now includes a multi-line list of hostname patterns (one per line). Auto-save runs only when the global toggle is on and the active tab matches a pattern. Defaults: Google Meet, Zoom, Microsoft Teams, Discord, and WhatsApp Web.
• Screen record during translation — New screen-record control next to the microphone records tab video and tab audio (microphone mixed in when the mic toggle is on). Stopping or ending the session downloads a WebM file. Free users: 10 minutes per recording; Pro: unlimited.
• Font settings dialog — Translation display font size, family, and color are configured from the context menu. The standalone font-size button was removed. Click Apply to update the live panel; settings persist in extension storage.
• Auto-save transcripts to Notion — Optional Auto Save to Notion toggle under Transcript settings. When enabled and you are signed in to Notion in Chrome, pick a workspace and database; live sessions sync transcript text to that database on the same interval as local auto-save (every 5 seconds and on stop). IndexedDB auto-save continues if Notion sync is off or fails.



v0.17.1 (May 7, 2026)

What's new for you

• ElevenLabs is now available as a narration provider — You can now add your own ElevenLabs API key in Settings and use ElevenLabs voices for both Narrate their voice and Narrate my voice.
• Inline AI Assistant is faster to use — Every assistant message now includes a one-click copy button, and hint buttons now append intelligently instead of overwriting your current input.
• Hint actions support double-click-to-send behavior — If a hint is already in your input, clicking that same hint again sends immediately, reducing repetitive editing steps.
• Cleaner selection helper behavior on websites — The selection assistant icon now appears after text selection is finalized (mouse release), reducing visual noise while dragging text.
• Selection assistant support is expanded — Selection-driven assistant entry points now cover more websites and keep page-specific selection context more stable.
• Keyboard shortcut status update — The planned assistant shortcut Ctrl+Shift+A / Cmd+Shift+A is not available yet in this version.



v0.14.7 (April 25, 2026)

What's new for you

• OpenAI warning is now shown only when needed — The extension no longer warns for missing OpenAI key at session start. You now see the warning only when using OpenAI-dependent actions (Narration with OpenAI voice, AI Assistant, and AI downloads).
• Cleaner state when YouTube caption narration is active — When narration switches to YouTube translated captions and Soniox is intentionally paused, the status now stays in Paused instead of showing a false Request timeout error.
• Close button now fully shuts down active voice features — Pressing the X closes the translator session and correctly stops active narration flows, including Narrate My Voice runtime state.
• Immediate original-volume recovery on close — If narration had ducked the original video volume, the extension now restores the original level before closing the translator panel.



v0.14.3 (April 23, 2026)

What's new for you

• Interpreter mode now works on Microsoft Teams chat — Teams CKEditor injection was upgraded so your translated speech can be inserted and sent reliably instead of failing with an empty-message warning.
• Narrate My Voice is now audible on Teams meetings — Teams voice routing was fixed so remote participants can hear your translated voice in real time.
• Clear guidance when Teams chat panel is closed — If Interpreter cannot send because Teams chat is not open, the popup now shows a user-facing warning instead of a console-only failure.
• New AI export: Download meeting summary — From the Download dialog, you can now generate and save a meeting summary as a .txt file.
• New AI export: Download task breakdown — You can now export action items with assignee details as .txt, and customize both summary/task prompts in OpenAI settings.



v0.13.1 (April 20, 2026)

What's new for you

• Browser local voice provider for Narrate their voice — You can now choose a Browser TTS provider and use local/browser-installed voices directly for speaker playback.
• Narrate My Voice keeps streaming injection path — My voice narration remains on OpenAI/Google streaming providers to preserve virtual mic injection reliability.
• Fixed long-queue sentence cutting — Browser narration queue handling was hardened to prevent skipped/truncated middle sentences in dense conversations.
• Reduced browser narration delay — Browser TTS batching/coalescing and provider-specific timing were tuned to lower repeated startup gaps between short segments.
• Faster and steadier speed behavior — Browser defaults are now optimized (1.75x default, 1.5x minimum), and auto-sync effective speed no longer drops below 1.5x.



v0.12.2 (April 19, 2026)

What's new for you

• YouTube caption-first narration for Their voice — On YouTube, Narrate their voice now prefers YouTube translated captions directly, so narration can start with lower translation cost and less delay.
• Safer fallback when captions are unavailable — If matching captions cannot be extracted, the extension now falls back automatically to the Soniox translation pipeline without breaking your narration flow.
• More stable timeline sync in YouTube narration — Caption narration now tracks start-at-offset playback and seek changes more accurately, while avoiding overlap between caption narration and Soniox narration.
• New Minimize action in translator header — You can now hide the panel using Minimize while keeping the extension runtime active in the background.
• Stop and Close now fully tear down runtime — Ending a session now consistently stops mic, narration, interpreter mode, and voice injection; pressing Stop from controls also auto-opens transcript download when data is available.
• Clear unsupported-page behavior for call actions — Microphone, interpreter, and Narrate My Voice actions now show disabled states with clearer tooltip feedback on unsupported pages.



v0.11.5 (April 11, 2026)

What's new for you

• Faster Narrate My Voice playback — Narrate My Voice now flushes earlier at punctuation and shorter buffer windows, so translated voice starts sooner during long sentences.
• Separate voice selection for each narration mode — You can now choose one voice for Narrate their voice and another voice for Narrate my voice, with each mode using exactly the selected voice at runtime.
• Voice preview for both voice selectors — Added a dedicated Test action for both Their voice and My voice so you can hear the selected voice before starting narration.
• Clearer narration controls — Added explanatory tooltips for narration actions and voice selectors, and moved Speaker settings into the control menu for a more consistent toolbar layout.
• Better first-time narration defaults — Default voice selections are now optimized for bilingual meetings: Their voice = Google Vietnamese, My voice = Google English.



v0.11.2 (April 10, 2026)

What's new for you

• New Context Settings for Soniox translation — You can now open context settings directly from the control bar via the control menu and define session context (domain, topic, terms, and speakers).
• Reusable saved contexts — Save context presets and quickly apply them in future sessions. Saved contexts now persist reliably after page reload and popup reopen.
• YouTube auto context mapping — On YouTube, the extension now auto-fills context using video metadata and auto-applies it for the current session (title -> domain, description -> topic).
• Import / Export context JSON — Export your saved contexts to JSON for backup, and import shared context packs from other users in one click.



v0.10.5 (April 9, 2026)

What's new for you

• Narrate My Voice now works on Zoom Web — Your translated voice is now delivered directly into Zoom Web meetings. Remote participants hear your speech in their language in real time.
• Interpreter chat injection for Zoom — Translated text can now be sent directly into Zoom's meeting chat, including support for Zoom's rich text editor.
• More reliable Narrate My Voice on WhatsApp — Fixed an issue where translated voice could stop working mid-call on WhatsApp Web.
• New platform: bringyourownlaptop.com — Live Translator now works on Bring Your Own Laptop courses and sessions.



v0.10.1 (April 6, 2026)

What's new for you

• New: WhatsApp Web support — Live Translator now works on WhatsApp Web, including real-time translation and Narrate My Voice for cross-language calls.
• Cleaner, more consistent controls — Start and End controls are now visually consistent and easier to use during fast conversations.
• Smarter session ending — Pressing End now fully stops active voice states: microphone, narration, and voice-injection mode.
• Better first-time narration defaults — Voice speed now starts at 60% and original volume at 20% for a more balanced listening experience.
• Easier setup guidance — Empty state messaging is clearer, and if OpenAI-dependent features are available but OpenAI key is missing, you'll see a direct warning with what to configure.



v0.9.3 (April 4, 2026)

What's new for you

• Auto update check when you open the popup — Live Translator now checks for a newer extension version each time you open the translator popup.
• In-app update badge — If a newer version is available, you'll see an Update available badge in the title bar. Click it to open the install page instantly.
• New subtitle mode for cleaner reading — Switch to a horizontal subtitle popup at the bottom-center of the screen to focus on translated captions without the full control-heavy layout.
• Subtitle customization — Subtitle mode now includes opacity and font-size controls so you can tune readability to your preference.
• Draggable and resizable subtitle view — Move and resize subtitle mode to match your meeting/video layout and screen space.



v0.8.5 (April 3, 2026)

What's new for you

• Echo-free meetings with extension enabled — Fixed the root echo issue so your calls stay clean even when the extension is installed and running.
• More stable Narrate My Voice pipeline — Improved virtual mic + translation routing behavior so continuous speech narration is more reliable.
• New narration mode flow — Narration Settings now uses two clear actions: Narrate my voice and Narrate their voice, with one-mode-at-a-time behavior.
• Simpler active state — When narration is active, the dialog now shows only Stop, making control easier and less confusing.
• Smarter narration status icon — The control icon now reflects mode state: voice_over_off (off), record_voice_over (my voice), transcribe (their voice).



v0.8.1 (April 2, 2026)

What's new for you

• Google + OpenAI in one voice picker — Narration settings now use one combined voice list, so you can switch engines by selecting a voice directly.
• Faster voice search — Voice selector now supports filtering/autocomplete for quick lookup by provider, language, or voice name.
• Google speed and volume controls improved — Voice speed and volume settings now apply properly when using Google voices.
• New platform support — Extension now supports running on www.domestika.org.



v0.7.4 (April 1, 2026)

What's new for you

• Lower narration latency — Narration now starts faster and responds quicker during continuous speech, so conversations feel more real time.
• Stable original volume while narrating — Original meeting/video audio now stays at your configured reduced level during narration playback instead of jumping up and down between sentences.
• Smart volume recovery — After narration stops, original volume stays reduced and only returns to normal if no new narration plays for 5 seconds.
• Smoother listening experience — Overall narration transitions are cleaner, with less distracting audio fluctuation.



v0.7.1 (March 31, 2026)

What's new for you

• Narrate My Voice is now more affordable — You can keep using translated voice in meetings with significantly lower API cost.



v0.6.5 (March 30, 2026)

What's new for you

• Faster chat send control — The Send my translation via chat action is now a quick button right in the main control bar (next to Microphone), so you can turn it on/off faster.
• Smoother interaction — Fixed the issue where this feature sometimes looked clickable but did not react in Settings.
• Clearer app info — The extension version is now shown in the title bar, so you can quickly verify what version you're using.



v0.6.2 (March 29, 2026)

What's new for you

• Narrate My Voice — Your translated speech is now spoken directly into the meeting. The other participant hears your words in their language in real time, so they don't need to read chat messages anymore.
• Smarter Narration — Narration now only reads translations from other participants and automatically skips your own voice. No more hearing your own words repeated back to you.
• SRT Subtitle Export — Download meeting transcripts as .srt subtitle files. Choose between original or translated text — ready to use in video editors or media players.
• Improved Controls — Microphone now defaults to off when you open the extension. Narration uses cleaner Start/Stop buttons and toggle switches for a simpler experience.
• Echo-Free Audio — Fixed audio feedback loop that could occur when narration and translation were both active. Conversations are now clean and echo-free.



v0.4.1 (March 24, 2026)

What's new for you

• Added Two-way Translation with My language and Their language selectors in the translator header.
• New Interpreter mode button: when enabled, your speech is translated into their language and can be sent directly to meeting chat.
• Language pair settings are now remembered automatically for future meetings.
• Improved Google Meet chat injection reliability with better input and send-button detection.



v0.3.1 (March 24, 2026)

What's new for you

• Fixed AI Assistant issue where your typed question could be missing in some requests during ongoing live translation.
• AI Assistant now uses one consistent assistant behavior, so responses are more predictable when you chat.
• You can now edit the General Assistant Prompt directly in OpenAI Settings (with a quick reset to default).



v0.3.0 (March 24, 2026)

What's new for you

• Inline AI Assistant is now built into each translation line, with quick actions to Explain and Generate answer.
• New dedicated Settings page: manage Soniox and OpenAI keys in one place with automatic save.
• Fresh, cleaner popup interface for better readability and easier control during live translation.
• Narration and Speaker settings dialogs are redesigned for a more consistent and simpler experience.



v0.2.0 (March 23, 2026)

What's new for you

• New Inline AI Assistant: hover any translation line and tap the assistant icon to get instant help.
• Two quick AI actions are now available: Explain (understand what the other person means) and Generate answer (get reply suggestions faster).



v0.1.1 (March 21, 2026)

• You can now change translation text size (Small, Medium, Large) directly from the control bar.
• Translation text size settings are saved, so your preferred view stays the same next time.
• New Narration mode is available: the app can read translated text out loud in real time.
• Narration now includes a simple settings panel (voice, speed, volume, and quick test).
• Playback is smoother and more reliable during live sessions, with better audio balance between narration and original video sound.