Cloud Speech 作者： Vivswan

Cloud Speech reads any highlighted text aloud with the cloud voice you choose. Bring your own API key for Amazon Polly, Azure Speech, Google Cloud Text-to-Speech, OpenAI, or any OpenAI-compatible server. The extension has no servers of its own: it talks only to the providers you give credentials to, your text goes only to the one whose voice you picked, and your keys are stored in your browser profile (synced through your browser account while Sync is on, the default), never sent to the extension's author.

HOW IT WORKS
1. Connect one or more providers in Settings with your own credentials. Save & test checks them.
2. Pick a voice in Preferences. Every voice from every connected provider is in one searchable picker, tagged by provider.
3. Highlight text on any page, then right-click > Read aloud, or press Ctrl+Shift+S (Command+Shift+S on Mac).

FEATURES
- Providers: Amazon Polly, Azure Speech, Google Cloud TTS, OpenAI, and OpenAI-compatible servers (Groq, DeepInfra, LiteLLM, LocalAI, Speaches, and others)
- One voice picker for all providers: search, filter by language, preview any voice before selecting it, star your favorites
- Read aloud from the context menu (normal, 1.5x, 2x speed), the keyboard shortcut, or the popup Sandbox
- Mini-player: play/pause, back and forward 15 seconds, live speed control
- Download any selection as an MP3 file: right-click > Download audio, or Ctrl+Shift+E (Command+Shift+E on Mac)
- Speed, pitch, volume gain, and speaking style, where the selected voice supports them
- Long selections are split into chunks automatically and stitched back together
- Interface in English, Hindi, Simplified Chinese, and Traditional Chinese
- Light, dark, or system theme
- Export and import your settings as a JSON file

YOUR KEYS, YOUR DATA
- Credentials are stored in your browser profile. While Sync is on (the default) the browser syncs them through your browser account; turn it off in Settings > Sync to keep them on this device only. They are never sent to the extension's author.
- Selected text is sent directly to the one provider whose voice you selected, with your own credentials.
- The other providers you gave credentials to are contacted only when you click Save & test (a short validation request and a voice check), when you preview one of their voices (a built-in sample sentence), and, while enabled, when their voice list is fetched (OpenAI's list is built in, and a voice list you type for an OpenAI-compatible server is used as is, so neither is asked). No other provider is contacted.
- The four named cloud providers are HTTPS-only; an OpenAI-compatible server URL you type yourself may be plain http, in which case your key and text travel unencrypted to that server.
- No analytics, no tracking, no servers of ours. The source code is public.
- Feedback > Report a bug opens a GitHub new-issue page in a new tab with the bug form's version, listing, environment, and provider fields prefilled from the URL (extension version, install source, browser with its version such as "Chrome 1xx...", and selected provider name), so GitHub sees them when the page opens. A value the extension does not know, such as the provider before you pick a voice, is left blank. You can edit everything before submitting.
- Feedback > Request a feature opens the feature form with nothing prefilled.
- Feedback > Leave a review opens this listing's review page on the store.
- Privacy policy: https://vivswan.github.io/cloud-speech/privacy/

PRICING
- The extension is free for individuals and for small organizations' internal use (Individual and Small Organization License, LICENSE.md in the repository). You pay your provider for what you use; most providers have a free tier.
- Comparison and tips: https://vivswan.github.io/cloud-speech/pricing/

SETUP GUIDES
- Amazon Polly: https://vivswan.github.io/cloud-speech/setup/polly/
- Azure Speech: https://vivswan.github.io/cloud-speech/setup/azure/
- Google Cloud TTS: https://vivswan.github.io/cloud-speech/setup/google/
- OpenAI: https://vivswan.github.io/cloud-speech/setup/openai/
- OpenAI-compatible services (Groq, DeepInfra, LiteLLM): https://vivswan.github.io/cloud-speech/setup/custom/
- Any other model (LiteLLM proxy, self-hosted LocalAI or Speaches): https://vivswan.github.io/cloud-speech/setup/local/

SUPPORT
- Troubleshooting: https://vivswan.github.io/cloud-speech/troubleshooting/
- Bugs and feature requests: https://github.com/vivswan/cloud-speech/issues
- Source code: https://github.com/vivswan/cloud-speech

FORMERLY POLLY FOR CHROME
Cloud Speech is the same extension, renamed. Amazon Polly is still fully supported, and settings from Polly for Chrome were carried over by the update.