Local AI Live Translate v1.9.1

What's new in 1.9.1

Easier pasting.

• API keys are saved as soon as you paste them. With Cloud AI providers on, pasting your key into the key field (Ctrl+V or right-click) saves it and loads your models straight away. Before, you had to click somewhere else first. The key is still never shown in the field.
• Right-click to paste into any text field in the popup (server addresses, API keys, names, search boxes), as in a terminal:
  • with text selected, right-click copies it instead;
  • Shift + right-click opens the browser's usual menu.
• New permission. This needs Chrome's "Read data you copy and paste" permission (clipboardRead), which the extension only uses when you right-click a text field.

To update, replace the old folder with this one (or run git pull on a clone), then reload the extension in chrome://extensions. If you've edited the templates in templates/, copy your versions across before replacing the folder.

Download

File	What it is
Local-AI-Live-Translate-<version>-windows-x64.zip	Everything, ready to run (Windows 10/11, 64-bit): server, speech models, a private Python and the Chrome extension. No installation and no internet needed to start it.
Local-AI-Live-Translate-extension-<version>.zip	The Chrome extension on its own (it is also inside the full zip).
SHA256SUMS.txt	Checksums to verify the downloads.

Get started

1. Unzip Local-AI-Live-Translate-<version>-windows-x64.zip anywhere and run START_Local_AI_Live_Translate.bat.
2. In Chrome, open chrome://extensions, turn on Developer mode, click Load unpacked and choose the extension folder.
3. Start LM Studio's local server with a translation model (Tencent Hy-MT2-7B recommended), or use Ollama or Qwen Cloud.
4. Open a video, click the extension icon, pick your languages and press Start Live Translate.

Running from a git clone instead? START_Local_AI_Live_Translate.bat downloads Python, its packages and the speech models on the first run (~1.1 GB, each file checked against a pinned checksum).

See the README for all the features and settings.

What's Changed

• Save pasted API keys straight away; right-click to paste in text fields by @xinjalin in #18

Full Changelog: v1.9.0...v1.9.1

Local AI Live Translate v1.9.0

What's new in 1.9.0

A new default look, matching the docs site.

• New Dark and Light themes (the default):

  • flat charcoal or white surfaces, with no glow;
  • a pale-blue status card;
  • squarer corners and panel titles in normal case instead of small capitals.

  They're the first themes that change the popup's layout as well as its colours. System picks them too.

• New logo colours. The toolbar icon matches: a pale-blue side with a dark "A", and a charcoal side with a pale-blue 文.

• The previous Dark and Light are still there, as Glass Dark and Glass Light.

• Hybrid has been removed. If you were using it, you'll get the new Dark; you can recreate it with Custom if you liked it.

• Custom themes have a new Studio style, with the layout of the new Dark and Light, alongside Glass and Flat.

• New docs site: guides, troubleshooting and a translation-model dashboard are now at xinjalin.github.io/Local-AI-Live-Translate-docs, and the README is much shorter.

If you were using the old Dark or Light theme, you'll see the new look after updating. Pick Glass Dark or Glass Light on the Display tab to go back.

To update, replace the old folder with this one (or run git pull on a clone), then reload the extension in chrome://extensions. If you've edited the templates in templates/, copy your versions across before replacing the folder.

Download

File	What it is
Local-AI-Live-Translate-<version>-windows-x64.zip	Everything, ready to run (Windows 10/11, 64-bit): server, speech models, a private Python and the Chrome extension. No installation and no internet needed to start it.
Local-AI-Live-Translate-extension-<version>.zip	The Chrome extension on its own (it is also inside the full zip).
SHA256SUMS.txt	Checksums to verify the downloads.

Get started

1. Unzip Local-AI-Live-Translate-<version>-windows-x64.zip anywhere and run START_Local_AI_Live_Translate.bat.
2. In Chrome, open chrome://extensions, turn on Developer mode, click Load unpacked and choose the extension folder.
3. Start LM Studio's local server with a translation model (Tencent Hy-MT2-7B recommended), or use Ollama or Qwen Cloud.
4. Open a video, click the extension icon, pick your languages and press Start Live Translate.

Running from a git clone instead? START_Local_AI_Live_Translate.bat downloads Python, its packages and the speech models on the first run (~1.1 GB, each file checked against a pinned checksum).

See the README for all the features and settings.

What's Changed

• Slim the README down and link to the new docs site by @xinjalin in #16
• New default Dark / Light theme with its own layout; Glass themes renamed, Hybrid removed by @xinjalin in #17

Full Changelog: v1.8.1...v1.9.0

Local AI Live Translate v1.8.1

What's new in 1.8.1

Fix: the Model menu for online providers can be opened.

• With Cloud AI providers on and an online provider picked, the Model field couldn't be clicked open once a model was filled in, so the list of your account's models couldn't be used. It's now a menu like the language menus:
  • click it to see your account's models;
  • search at the top to find one;
  • pick with the mouse, or with the arrow keys and Enter.
• A model that isn't in the list: type its name in the search box and pick Use "…".

To update, replace the old folder with this one (or run git pull on a clone), then reload the extension in chrome://extensions. If you've edited the templates in templates/, copy your versions across before replacing the folder.

Download

File	What it is
Local-AI-Live-Translate-<version>-windows-x64.zip	Everything, ready to run (Windows 10/11, 64-bit): server, speech models, a private Python and the Chrome extension. No installation and no internet needed to start it.
Local-AI-Live-Translate-extension-<version>.zip	The Chrome extension on its own (it is also inside the full zip).
SHA256SUMS.txt	Checksums to verify the downloads.

Get started

1. Unzip Local-AI-Live-Translate-<version>-windows-x64.zip anywhere and run START_Local_AI_Live_Translate.bat.
2. In Chrome, open chrome://extensions, turn on Developer mode, click Load unpacked and choose the extension folder.
3. Start LM Studio's local server with a translation model (Tencent Hy-MT2-7B recommended), or use Ollama or Qwen Cloud.
4. Open a video, click the extension icon, pick your languages and press Start Live Translate.

Running from a git clone instead? START_Local_AI_Live_Translate.bat downloads Python, its packages and the speech models on the first run (~1.1 GB, each file checked against a pinned checksum).

See the README for all the features and settings.

What's Changed

• Fix: online provider Model field couldn't be opened (now a searchable menu) by @xinjalin in #15

Full Changelog: v1.8.0...v1.8.1

Local AI Live Translate v1.8.0

What's new in 1.8.0

Online AI providers, only if you want them, and safer API keys.

• Nothing goes online unless you choose it. The old "cloud translation backup" is gone. Before, when your local model couldn't be reached, a line could be sent to DeepSeek (if you'd entered a key) and then to Google Translate, which needed no key and was on by default. Now a line the local model can't translate is shown untranslated.
• Cloud AI providers (Model tab, off by default). Turning it on first explains what it means: while an online provider is your LLM server, the text of everything captured is sent to it, handled under its privacy terms, and billed to your API key.
  • While it's on, LLM Server also lists Qwen Cloud, OpenAI, Anthropic Claude, DeepSeek, Google Gemini and xAI Grok. Pick one, paste your key (the Get a key link opens the provider's key page) and choose a model from your account's list, or type one.
  • Turning it off hides them again and switches back to LM Studio.
  • The footer reads Cloud AI providers on while the setting is on.
• Safer API keys:
  • Where they're kept: the extension's own private storage. Web pages can't read it, and neither can the extension's script inside them. They used to be kept where that script could see them.
  • Where they're sent: only to their own provider's official HTTPS address, never anywhere else.
  • What the popup shows: never a saved key, only its last four characters. Keys are never in profiles, exports, transcripts or the server's log.
  • Saved API Keys (Model tab) lists them: delete one, or Clear all API keys. Keys saved by older versions are moved there automatically.
• Web pages can't use the app server. It now only accepts the extension, so a site open in your browser can't connect to it or read what it reports.
• Profiles remember the online provider and model, never the key. A profile that uses an online provider falls back to LM Studio while cloud providers are off, and says why.

Updating:

• If you used Qwen Cloud: it stays selected, and Cloud AI providers starts switched on.
• If you entered a DeepSeek key for the old backup: it's now listed under Saved API Keys. It's only used if you pick DeepSeek as your LLM server; delete it there if you don't want it.
• Keys are stored like any browser extension data, not encrypted. On a shared PC, use Clear all API keys when you're done.

To update, replace the old folder with this one (or run git pull on a clone), then reload the extension in chrome://extensions. If you've edited the templates in templates/, copy your versions across before replacing the folder.

Download

File	What it is
Local-AI-Live-Translate-<version>-windows-x64.zip	Everything, ready to run (Windows 10/11, 64-bit): server, speech models, a private Python and the Chrome extension. No installation and no internet needed to start it.
Local-AI-Live-Translate-extension-<version>.zip	The Chrome extension on its own (it is also inside the full zip).
SHA256SUMS.txt	Checksums to verify the downloads.

Get started

1. Unzip Local-AI-Live-Translate-<version>-windows-x64.zip anywhere and run START_Local_AI_Live_Translate.bat.
2. In Chrome, open chrome://extensions, turn on Developer mode, click Load unpacked and choose the extension folder.
3. Start LM Studio's local server with a translation model (Tencent Hy-MT2-7B recommended), or use Ollama or Qwen Cloud.
4. Open a video, click the extension icon, pick your languages and press Start Live Translate.

Running from a git clone instead? START_Local_AI_Live_Translate.bat downloads Python, its packages and the speech models on the first run (~1.1 GB, each file checked against a pinned checksum).

See the README for all the features and settings.

What's Changed

• README: add Hy-MT2-1.8B and Hy-MT2-7B BF16 benchmark results by @xinjalin in #13
• Optional cloud AI providers with safer API key handling; remove the online fallback by @xinjalin in #14

Full Changelog: v1.7.0...v1.8.0

Local AI Live Translate v1.7.0

What's new in 1.7.0

Two cherry-blossom themes.

• Sakura Light and Sakura Dark (Display tab → Theme). Both use the Glass style, with a soft pink glow and see-through panels:
  • Sakura Light: petal pinks on a pale blush background, with a deep sakura button.
  • Sakura Dark: blossom pinks on a deep plum background, with a pale petal button.
• Start a Custom theme from Sakura. Both are in the Custom theme's Start From list, so you can use them as a base for your own colors.
• Light Custom themes look closer to the built-in Light theme. In a light Glass theme, the status card is now tinted with your accent color instead of fading to grey.

Like the other themes, Sakura can be saved with a display config and is carried across by export and import.

To update, replace the old folder with this one (or run git pull on a clone), then reload the extension in chrome://extensions. If you've edited the templates in templates/, copy your versions across before replacing the folder.

Download

File	What it is
Local-AI-Live-Translate-<version>-windows-x64.zip	Everything, ready to run (Windows 10/11, 64-bit): server, speech models, a private Python and the Chrome extension. No installation and no internet needed to start it.
Local-AI-Live-Translate-extension-<version>.zip	The Chrome extension on its own (it is also inside the full zip).
SHA256SUMS.txt	Checksums to verify the downloads.

Get started

1. Unzip Local-AI-Live-Translate-<version>-windows-x64.zip anywhere and run START_Local_AI_Live_Translate.bat.
2. In Chrome, open chrome://extensions, turn on Developer mode, click Load unpacked and choose the extension folder.
3. Start LM Studio's local server with a translation model (Tencent Hy-MT2-7B recommended), or use Ollama or Qwen Cloud.
4. Open a video, click the extension icon, pick your languages and press Start Live Translate.

Running from a git clone instead? START_Local_AI_Live_Translate.bat downloads Python, its packages and the speech models on the first run (~1.1 GB, each file checked against a pinned checksum).

See the README for all the features and settings.

What's Changed

• Add Sakura Light and Sakura Dark themes by @xinjalin in #12

Full Changelog: v1.6.0...v1.7.0

Local AI Live Translate v1.6.0

What's new in 1.6.0

Make the popup your own with a Custom theme.

• Custom theme (Display tab → Theme → Custom). Choose:

  • Style: Glass (glowing background and see-through panels, like Dark and Light) or Flat (solid surfaces, like the OLED themes).
  • Start From: any built-in theme, as a starting point. The first time you choose Custom, it starts from the theme you had on.
  • Colors: background, panels, accent and buttons.
  • Text: headings, labels, text, hints and dropdowns.

  Everything else is worked out from those colors. If a text color would be hard to read on your background, it's lightened or darkened just enough to stay readable. Changes show instantly while you drag a color picker, and nothing is sent to the server.

• Themes are saved with display configs. A display config now keeps the theme too, with your Custom colors when that's the theme in use, so you can keep several themes and switch between them. This way there's still only two kinds of config: profiles for the model and languages, and display configs for how things look. Display configs saved before 1.6.0 don't include a theme and leave yours as it is. Export and import carry the theme across, and imported colors are checked.

• Updated screenshots in the README, taken from this version.

To update, replace the old folder with this one (or run git pull on a clone), then reload the extension in chrome://extensions. If you've edited the templates in templates/, copy your versions across before replacing the folder.

Download

File	What it is
Local-AI-Live-Translate-<version>-windows-x64.zip	Everything, ready to run (Windows 10/11, 64-bit): server, speech models, a private Python and the Chrome extension. No installation and no internet needed to start it.
Local-AI-Live-Translate-extension-<version>.zip	The Chrome extension on its own (it is also inside the full zip).
SHA256SUMS.txt	Checksums to verify the downloads.

Get started

1. Unzip Local-AI-Live-Translate-<version>-windows-x64.zip anywhere and run START_Local_AI_Live_Translate.bat.
2. In Chrome, open chrome://extensions, turn on Developer mode, click Load unpacked and choose the extension folder.
3. Start LM Studio's local server with a translation model (Tencent Hy-MT2-7B recommended), or use Ollama or Qwen Cloud.
4. Open a video, click the extension icon, pick your languages and press Start Live Translate.

Running from a git clone instead? START_Local_AI_Live_Translate.bat downloads Python, its packages and the speech models on the first run (~1.1 GB, each file checked against a pinned checksum).

See the README for all the features and settings.

What's Changed

• Add a Custom popup theme, saved with display configs by @xinjalin in #11

Full Changelog: v1.5.0...v1.6.0

Local AI Live Translate v1.5.0

What's new in 1.5.0

Prompt templates, so more translation models work well.

• Prompt Template menu (Model tab). A template decides how each subtitle line is put to the model. Auto picks one from the model's name:

  • MiLMMT (Xiaomi) gets its own "Translate this from X to Y" format. Xiaomi's MiLMMT-46 models translate properly in the app now; before this they returned nothing.
  • Hy-MT (Tencent) gets Tencent's official template, as before.
  • General covers every other model.

  The choice is saved with profiles and included in export and import.

• Your own templates. The app's templates are editable JSON files in the new templates/ folder (generic.json, hy-mt.json, milmmt.json). Add your own there: a prompt with {text}, {target}, {source} and {context} placeholders, match words for Auto, and sampling settings. See templates/README.md for the format, and copy _example.json to start. If a file is missing or broken, the built-in copy is used, so translation keeps working.

• Chinese subtitles always in the script you chose. Traditional or Simplified: stray characters are converted whatever the model writes.

• Find and download for imported profiles. When an imported profile's model isn't installed and its download source is unknown (for example, it was exported while the app server wasn't running), Find and download looks the model up on Hugging Face by name, publisher and quantization, and LM Studio downloads it. This only happens when you press the button, and only the model's name and publisher are sent. Exports now also record each model's publisher.

• Clearer delete confirmation. The first press on a profile's or display config's delete button now shows "Press 🗑 again to delete …" with a countdown line that shrinks over the 4 seconds you have to confirm.

To update, replace the old folder with this one (or run git pull on a clone), then reload the extension in chrome://extensions. If you've edited the templates in templates/, copy your versions across before replacing the folder.

Download

File	What it is
Local-AI-Live-Translate-<version>-windows-x64.zip	Everything, ready to run (Windows 10/11, 64-bit): server, speech models, a private Python and the Chrome extension. No installation and no internet needed to start it.
Local-AI-Live-Translate-extension-<version>.zip	The Chrome extension on its own (it is also inside the full zip).
SHA256SUMS.txt	Checksums to verify the downloads.

Get started

1. Unzip Local-AI-Live-Translate-<version>-windows-x64.zip anywhere and run START_Local_AI_Live_Translate.bat.
2. In Chrome, open chrome://extensions, turn on Developer mode, click Load unpacked and choose the extension folder.
3. Start LM Studio's local server with a translation model (Tencent Hy-MT2-7B recommended), or use Ollama or Qwen Cloud.
4. Open a video, click the extension icon, pick your languages and press Start Live Translate.

Running from a git clone instead? START_Local_AI_Live_Translate.bat downloads Python, its packages and the speech models on the first run (~1.1 GB, each file checked against a pinned checksum).

See the README for all the features and settings.

What's Changed

• Find and download a profile's model when its download source is unknown by @xinjalin in #9
• Prompt templates (MiLMMT now translates), templates folder, clearer delete confirmation by @xinjalin in #10

Full Changelog: v1.4.0...v1.5.0

Local AI Live Translate v1.4.0

What's new in 1.4.0

A roomier popup that fits every UI language, and language menus you can search.

• Searchable language menus. The video, subtitle and UI language menus have a search box pinned to the top of the list. It matches exact text, letters in order and small typos ("japnese" finds Japanese). It also recognises each language's English name, its own name (日本 finds Japanese) and common aliases such as Bengali, Tagalog, 粵語 and Mandarin. Languages are listed A–Z in the popup's own language, and you can open a menu by pressing Enter or by starting to type.
• Consistent, compact dropdowns. Every other dropdown now opens the same rounded menu as the language menus, with a check mark on the current choice, in all seven themes. The font menu shows each font in its own typeface. This uses Chrome's customisable select, so it needs Chrome 135 or newer; older versions keep the plain system list.
• A wider popup: 400 px, up from 360. Nothing wraps, gets cut off or shifts in any of the 10 UI languages:
  • The UI Language label now has a globe icon instead of an English "(UI Language)" suffix.
  • The Profile and Display Config buttons moved up beside their titles, so those dropdowns are full width.
  • The German, Spanish, French and Russian taglines are shorter and closer to the English.
• README: new translation model comparison. Hy-MT2-7B, Hy-MT2-30B-A3B, MiLMMT-46 (12B and 4B), Gemma-4-12B and Qwen3.5-9B, measured across 39 language pairs. Hy-MT2-7B is still the recommended model.

To update, replace the old folder with this one (or run git pull on a clone), then reload the extension in chrome://extensions.

Download

File	What it is
Local-AI-Live-Translate-<version>-windows-x64.zip	Everything, ready to run (Windows 10/11, 64-bit): server, speech models, a private Python and the Chrome extension. No installation and no internet needed to start it.
Local-AI-Live-Translate-extension-<version>.zip	The Chrome extension on its own (it is also inside the full zip).
SHA256SUMS.txt	Checksums to verify the downloads.

Get started

1. Unzip Local-AI-Live-Translate-<version>-windows-x64.zip anywhere and run START_Local_AI_Live_Translate.bat.
2. In Chrome, open chrome://extensions, turn on Developer mode, click Load unpacked and choose the extension folder.
3. Start LM Studio's local server with a translation model (Tencent Hy-MT2-7B recommended), or use Ollama or Qwen Cloud.
4. Open a video, click the extension icon, pick your languages and press Start Live Translate.

Running from a git clone instead? START_Local_AI_Live_Translate.bat downloads Python, its packages and the speech models on the first run (~1.1 GB, each file checked against a pinned checksum).

See the README for all the features and settings.

What's Changed

• Searchable language menus, rounded dropdowns everywhere, model benchmark in README by @xinjalin in #7
• Wider popup so every UI language fits; consistent compact dropdowns by @xinjalin in #8

Full Changelog: v1.3.0...v1.4.0

Local AI Live Translate v1.3.0

What's new in 1.3.0

Eight more languages. Portuguese, Italian, Turkish, Polish, Ukrainian, Dutch and Bengali can now be used for both the video and the subtitles. Cantonese can be used as a video language.

Language	Speech engine	Errors in tests (Google FLEURS)
Cantonese	SenseVoice	4.7% of characters
Portuguese	Omnilingual	14.3% of words (4.0% of characters)
Italian	Omnilingual	13.4% (1.7%)
Dutch	Omnilingual	22.7% (7.1%)
Ukrainian	Omnilingual	25.7% (4.6%)
Polish	Omnilingual	24.9% (4.1%)
Turkish	Omnilingual	22.2% (4.0%)
Bengali	Omnilingual	15.5% (7.2%)

• The best engine is picked automatically. If you choose one that can't recognise the language, the popup says which one it switched to. Omnilingual is the default for the European languages and Turkish because it's 3–6× faster. Whisper-Small is a little more accurate for Italian and Dutch, but adds 1–2 s per line, and you can choose it on the Live tab.
• Cantonese is shown in Traditional characters. Chinese subtitles for Cantonese are now translated into standard written Chinese (呢幾個字都表達唔到 → 這幾個字都表達不了), where before the characters were only converted. This also applies when Cantonese is auto-detected, and adds about 0.3–0.5 s per line. Cantonese isn't offered as a subtitle language, because the translation model writes formal Chinese rather than colloquial Cantonese.
• Hy-MT2-7B is still the recommended translator. In a new 39-language-pair comparison, it matched Xiaomi MiLMMT-46-12B and Tencent's Hy-MT2-30B-A3B for quality and was the fastest of the three (485 ms per line).
• Known limitations:
  • The original-language line from Omnilingual has no punctuation or capitals.
  • Urdu isn't included yet, because no engine recognises it well in Urdu script.

There are no new downloads: the new languages use the speech models already included. To update, replace the old folder with this one, or run git pull on a clone, then reload the extension in chrome://extensions.

Download

File	What it is
Local-AI-Live-Translate-<version>-windows-x64.zip	Everything, ready to run (Windows 10/11, 64-bit): server, speech models, a private Python and the Chrome extension. No installation and no internet needed to start it.
Local-AI-Live-Translate-extension-<version>.zip	The Chrome extension on its own (it is also inside the full zip).
SHA256SUMS.txt	Checksums to verify the downloads.

Get started

1. Unzip Local-AI-Live-Translate-<version>-windows-x64.zip anywhere and run START_Local_AI_Live_Translate.bat.
2. In Chrome, open chrome://extensions, turn on Developer mode, click Load unpacked and choose the extension folder.
3. Start LM Studio's local server with a translation model (Tencent Hy-MT2-7B recommended), or use Ollama or Qwen Cloud.
4. Open a video, click the extension icon, pick your languages and press Start Live Translate.

Running from a git clone instead? START_Local_AI_Live_Translate.bat downloads Python, its packages and the speech models on the first run (~1.1 GB, each file checked against a pinned checksum).

See the README for all the features and settings.

What's Changed

• Credit LiveCaption by begin0808 as the project's origin by @xinjalin in #4
• Add a Screenshots section to the README by @xinjalin in #5
• Add Cantonese, Portuguese, Italian, Turkish, Polish, Ukrainian, Dutch and Bengali by @xinjalin in #6

Full Changelog: v1.2.1...v1.3.0

Local AI Live Translate v1.2.1

What's new in 1.2.1

Fixed: lag when dragging a colour picker or slider. Every step of a drag used to send the whole config to the translation server and redraw the subtitles on the page, dozens of times a second.

• Subtitle look (colours, opacity, outline and so on): the preview in the popup follows every step. The subtitles on the page update up to 10 times a second while you drag, and once more when you close the picker or let go. None of this goes to the server any more; it never used these settings.
• Server settings (languages, model, speech tuning, API keys): sent only once they've changed and stayed the same for 0.3 s. Dragging a slider sends a single update, and typing a Qwen Cloud API key no longer restarts the connection on every keystroke.
• Server log: a repeat of the same config is no longer logged again.

To update, replace the old folder with this one, or run git pull on a clone, then reload the extension in chrome://extensions.

Download

File	What it is
Local-AI-Live-Translate-<version>-windows-x64.zip	Everything, ready to run (Windows 10/11, 64-bit): server, speech models, a private Python and the Chrome extension. No installation and no internet needed to start it.
Local-AI-Live-Translate-extension-<version>.zip	The Chrome extension on its own (it is also inside the full zip).
SHA256SUMS.txt	Checksums to verify the downloads.

Get started

1. Unzip Local-AI-Live-Translate-<version>-windows-x64.zip anywhere and run START_Local_AI_Live_Translate.bat.
2. In Chrome, open chrome://extensions, turn on Developer mode, click Load unpacked and choose the extension folder.
3. Start LM Studio's local server with a translation model (Tencent Hy-MT2-7B recommended), or use Ollama or Qwen Cloud.
4. Open a video, click the extension icon, pick your languages and press Start Live Translate.

Running from a git clone instead? START_Local_AI_Live_Translate.bat downloads Python, its packages and the speech models on the first run (~1.1 GB, each file checked against a pinned checksum).

See the README for all the features and settings.

What's Changed

• Stop colour and slider drags from flooding the server with config updates by @xinjalin in #3

Full Changelog: v1.2.0...v1.2.1