Media
3 items
Description
A voice and text browser agent that researches, navigates, fills, compares, and verifies—with safe checkpoints and advanced Arabic. Thahen AI (Work) turns a typed or spoken request into a clear task inside Chrome. Your request, live work, questions, approvals, and result appear in one side conversation. Thahen explains what it is doing and where it has reached. You choose the execution permission at the bottom of the conversation. What it can do: - Search within websites, navigate results, and gather relevant information. - Open task tabs, compare options, and produce evidence-backed recommendations. - Fill fields, choose values, click, and scroll, while pausing before submission or sensitive action. - Work naturally across Arabic, English, and mixed-language pages, with strong Saudi and Gulf context. - Use bounded visual context when needed for an image, PDF, video, or low-text page. - Summarize a full YouTube video from available timed captions, or listen directly to the playing video with bounded visual frames when captions are unavailable. - Explicitly attach a PDF, Word, Excel, PowerPoint, or image to a task and compare it with website information. - Resume multi-step work after the panel closes, with clear pause, resume, and stop controls. - Hold a continuous voice task conversation: talk to the agent, hear progress and questions, interrupt naturally, and redirect the active task while it is working. - Show your messages, completed steps, current plan, usage, evidence, and result in the Chrome Side Panel. Choose your control level: - **Ask before each action:** review each browser action before it runs. - **Automatic execution:** let routine actions run within the stated goal. - **Automatic on this site:** work smoothly on the site you permitted. - **Read only:** research and extract without changing the page or filling fields. Safety is part of execution. Thahen asks once before sending, posting, uploading, downloading, or recoverable deletion. It never reads or fills passwords, OTP/CAPTCHA codes, and never completes a payment, purchase, legal acceptance, final publish, or irreversible deletion. It hands those steps to you, then resumes. Thahen also includes a fast content assistant for text you choose: executive briefs, source comparison, decision memos, claims review, table extraction, summarization, translation, rewriting, proofreading, and drafting. Results are rendered as safe text; generated HTML or JavaScript is never executed. Thahen does not read browsing history or unrelated tabs. Website access is requested when a task needs it, and you can deny or revoke it in Chrome. There is no remotely hosted executable code, advertising use, or browsing tracking. Thahen never chooses files from your device. You explicitly select a task file and can remove it before starting; it is prepared within the request and no copy is written to server storage in this flow. Chrome protects internal browser pages and the Chrome Web Store itself from every extension. Thahen explains this in the conversation and keeps your request ready until you open a regular website. The microphone never starts ambiently. You explicitly start a visible, low-latency conversation that remains active only while that conversation is on, allowing natural speech, spoken progress, and interruption. No audio copy is written to server file storage. You can press **End**, mute spoken responses, or revoke microphone access at any time. Video understanding never starts automatically. You explicitly press the video button. Thahen first uses timed captions made available to the YouTube page; otherwise it captures the player audio directly, with current-tab audio as a compatibility fallback, plus a bounded set of visible frames until the video ends or you stop it. The extension uses your existing unified Thahen account, plan, permissions, and balance. Account and plan management remain on the Thahen website; there is no separate extension subscription.