Item logo image for Telli.sh: Every Translator — Read, Reply & Summarize Any Site

Telli.sh: Every Translator — Read, Reply & Summarize Any Site

ExtensionCommunication19 users
Item media 5 (screenshot) for Telli.sh: Every Translator — Read, Reply & Summarize Any Site
Item media 1 (screenshot) for Telli.sh: Every Translator — Read, Reply & Summarize Any Site
Item media 2 (screenshot) for Telli.sh: Every Translator — Read, Reply & Summarize Any Site
Item media 3 (screenshot) for Telli.sh: Every Translator — Read, Reply & Summarize Any Site
Item media 4 (screenshot) for Telli.sh: Every Translator — Read, Reply & Summarize Any Site
Item media 5 (screenshot) for Telli.sh: Every Translator — Read, Reply & Summarize Any Site
Item media 1 (screenshot) for Telli.sh: Every Translator — Read, Reply & Summarize Any Site
Item media 1 (screenshot) for Telli.sh: Every Translator — Read, Reply & Summarize Any Site
Item media 2 (screenshot) for Telli.sh: Every Translator — Read, Reply & Summarize Any Site
Item media 3 (screenshot) for Telli.sh: Every Translator — Read, Reply & Summarize Any Site
Item media 4 (screenshot) for Telli.sh: Every Translator — Read, Reply & Summarize Any Site
Item media 5 (screenshot) for Telli.sh: Every Translator — Read, Reply & Summarize Any Site

Overview

Read translations next to the original, reply in any language, and right-click any page into a summarized Telli.sh note.

Telli.sh is a browser extension that helps you read, reply and keep useful material from the web. Inline translation sits beside the original text, a writing panel helps you reply in another language, and a side panel lets you discuss the page you are viewing. Read a summary to understand the topic, examine the key points, then save the page and its conversation together when you choose. The original text remains available for comparison throughout your reading and writing. READ IN CONTEXT: 4 DISPLAY MODES - Inline comment: place the result below the original message, with a separate visual treatment that makes the two easy to distinguish. Follow a discussion without moving its content to another tab. - Bilingual: keep both versions visible together. Compare a term, check a sentence or return to a detail in the source while you read the surrounding discussion. - Click to expand: show a compact control until you want the full result. Open it when needed and collapse it again to keep a busy page manageable. - Click to toggle: switch between the original and the result in the same area. Choose this mode when space matters and you prefer to inspect one version at a time. - On supported pages, newly loaded messages and comments are detected as you scroll. An item's refresh control lets you request a new result after a message expands or its source changes. Emoji-only messages, bare links and text recognized as already in your preferred language can be skipped. WRITE REPLIES WITH A PREVIEW - Draft a response in your own language in the floating writing panel. Read the version intended for the recipient before you put it into the page. This is useful for everyday chats, community discussions and longer replies where wording matters. - With send preview enabled, edit the prepared result before continuing. Check a name, adjust a greeting or keep your preferred terminology. You decide when the reviewed text is ready to use. - On supported conversation pages, the outgoing language can follow the other person's messages. Configure a fallback for a new conversation where there is not enough context to infer the recipient's language yet. - In insertion-only editors, including supported mail composers, the browser extension fills the field and leaves the final send action to you. Website editors behave differently, so the available action follows the integration for that page. - Multiline drafts keep their line breaks through processing and insertion. Separate paragraphs and short lists can remain part of your message. INPUT-FIELD TRANSLATION ON ADDITIONAL SITES - Turn on the optional input assistant to use text fields on ordinary web pages beyond the supported-site catalogue. This feature is disabled by default. When you enable it, the browser extension requests the site access it needs. - A small language control appears beside a supported field when you focus it. Use it while writing a form response, comment or message without copying the entire draft into a separate service. - Let the target follow the site's language or select a fixed language in Settings. Exclude individual domains when you want the assistant available elsewhere but quiet on a particular site. - Compatibility depends on the website's text editor. Protected browser pages and inaccessible text fields are not supported. POPUP TOOLS AND TRANSLATION HISTORY - The popup retains three tabs: a text workspace, history and status. Paste a passage into the workspace, select source and target languages, and work with up to 2,000 characters at a time. Automatic source detection is available when you do not want to choose it yourself. - Revisit previous popup results in the history tab. Copy useful wording, pin an entry you want to keep, or delete an item you no longer need. These entries are stored on this device, separately from saved web clips. - Unpinned history entries expire after 30 days. Pinning keeps an item beyond that routine expiry, within the local history capacity. This makes it easier to retain a useful phrase without keeping every temporary lookup indefinitely. - The status tab brings language settings, display mode, send preview, the selected engine and the input assistant into one view. A small activity indicator in the header shows the page's enabled state, with site information on hover. SIDE PANEL: PAGE CHAT, SUMMARIES AND KEY POINTS - Keep the page visible while asking questions about it. The option to use the current page controls whether its text is included as context. That text is read when you request an action with the option enabled. - Ask for an overview to understand the main argument, extract key points to find the central claims, or request an explanation of a passage. Follow up in the same conversation to narrow the question or examine a detail. - The page actions include a language conversion, a summary, key points and a source-text view. Inspect the extracted text when you want to understand what material is available to the assistant, and compare an answer with the actual page. - Actions reflect the selected engine's capabilities and the availability of page context. A language-only service can handle text conversion; page questions and summaries need a conversation-capable engine. When a required capability is unavailable, the panel explains how to change the engine in Settings. - Choose the engine and languages in Settings. Each result identifies the responding provider so you can see which configured engine answered. - Start a new conversation from the header, return to earlier conversations from the list, or delete conversations there. Press Enter to send your question. Press Shift+Enter to insert a line break. Switch between popup and side-panel presentation according to the space you need. VIDEO SUBTITLE SUMMARIES - On supported video pages, the side panel can use available video subtitles or the page's transcript to explain the spoken content. Summarize a presentation, extract the key points of a tutorial or ask about an idea in the available transcript while keeping the video page open. - When a transcript is long, its text can be processed in sections and combined into a final overview. Available time ranges help preserve the order of the discussion. The interface reports when source limits shorten the material being used. - This feature works from available text. It does not listen to audio, inspect video frames, generate missing subtitles or unlock material that the page does not expose. If video subtitles cannot be read, the panel may use the accessible page text instead; a description alone is not the full spoken content. - View the extracted text to check the available material before relying on a summary. Video subtitles may contain mistakes or omit part of a talk, so the completeness of the answer depends on what the page makes available and what the selected model can process. - The reading integration also handles descriptions and comments on supported video pages. These page elements are distinct from the video's subtitle transcript. 52 LANGUAGE OPTIONS AND 50 LOCALIZED INTERFACES - Choose from 52 configured target-language options, or let the target follow your system language. Incoming text and outgoing replies can use different targets, so reading comfortably does not mean asking everyone else to use your language. - The interface ships in 50 language and regional localizations. Settings, status messages and everyday controls follow the available interface locale, while the language of the material you read or write is a separate choice. - Coverage depends on the selected engine and language pair. The presence of a language in the selector does not mean that every local model or remote service supports it equally. Check readiness in Settings and choose an engine suited to the languages you need. CHOOSE HOW TEXT IS PROCESSED - Start with the default online service without creating an account or supplying an API key for basic use. Configure an alternative when you need a different language range, a particular model or control over where requests go. - Available approaches include browser-integrated on-device processing, a local model server, key-based cloud services and compatible API endpoints. Provider details and connection controls live in Settings, where you can prepare a supported model or check the configured connection. - Dedicated translation engines and conversation models serve different needs. The first category focuses on converting text between languages; the second also supports questions, summaries and key-point extraction. Select a conversation-capable option to use those additional panel actions. - Built-in AI depends on browser support, device resources, model preparation and language availability. Local-server use requires a running, reachable endpoint. Cloud options may need your own credentials and may have separate usage limits or charges set by their provider. - Supported model selectors retrieve their catalogue from the provider and show pricing when available. Consumer subscriptions do not automatically grant API access. CLIP: SAVE A PAGE AND ITS CONVERSATION TO Telli.sh - Connect a Clip token in Settings to save material to your account. Use the page's right-click action to keep its URL, available source metadata and selected text, or use the conversation's Clip control to keep the page together with its discussion. - A conversation clip includes the exchange you choose to save, rather than requiring you to save each answer separately. The source address travels with it, so you can later return to the page that prompted the questions. - Available source information can include the page title, canonical address, description, preview image, site name, author and publication date. What is available depends on the page; a clip does not automatically include the full article body or everything behind the link. - AI-generated titles and summaries depend on the submitted content and your account allowance. They describe the available material, which may cover only part of a page. The interface reports the applicable result when an allowance is exhausted. - Saving is an explicit action. Browsing alone does not create a collection of clips, and a local popup history entry is not automatically uploaded. You decide which useful pages and conversations become part of your knowledge collection. - Clip access stays off until a token is connected. The token is limited to saving clips, is stored on this device and does not grant access to account billing. Remove it in Settings to disconnect this saving path. SITE SETTINGS AND EVERYDAY CONTROL - Browse the supported sites by category, including messaging, mail, communities, reference material, articles and comments. Enable or disable the integration for a site without changing how you use the others. A request link is available for missing sites. - Configure the languages, display mode, preview-before-sending option and provider on the settings page. Changes are saved immediately. Visual examples show how the chosen display works before you return to a conversation. PRIVACY AND PROCESSING BOUNDARIES - Page conversations use the selected provider. Some conversion paths retry with the default online service when the chosen engine fails or an on-device language pair is unavailable. Results identify the engine that responded. Browser on-device conversation does not fall back to a network provider. - API keys and Clip tokens remain in local extension storage rather than being synced through your browser account. The browser extension does not send usage analytics or telemetry. Remote engines receive the text needed for the requests they process under their own service terms. - Page context is read for the actions you request, and only clips you explicitly save are sent to the connected knowledge service. Optional site permissions are requested when the enabled feature needs them. The browser extension does not bypass sign-in requirements, paywalls or private-content controls. Feature guide: https://telli.sh/sns-translator-extension Privacy: https://telli.sh/sns-translator-extension/privacy-policy

Details

  • Version
    1.29.6
  • Updated
    September 13, 2026
  • Size
    543KiB
  • Languages
    50 languages
  • Developer
    Website
    Email
    jo.dreame@gmail.com
  • Non-trader
    This developer has not identified itself as a trader. For consumers in the European Union, please note that consumer rights do not apply to contracts between you and this developer.

Privacy

Manage extensions and learn how they're being used in your organization

Telli.sh: Every Translator — Read, Reply & Summarize Any Site has disclosed the following information regarding the collection and usage of your data. More detailed information can be found in the developer's privacy policy.

Telli.sh: Every Translator — Read, Reply & Summarize Any Site handles the following:

Authentication information
Personal communications
Website content

This developer declares that your data is

  • Not being sold to third parties, outside of the approved use cases
  • Not being used or transferred for purposes that are unrelated to the item's core functionality
  • Not being used or transferred to determine creditworthiness or for lending purposes
Google apps