Doitong Copilot

Doitong Copilot — doitong.com 的 AI 助手。透過與瀏覽器對話，建立故事板、生成影片和圖像。

Doitong Copilot is your sidebar-resident AI creative studio. Talk to it. Type at it. Tell it to make pictures and videos. It's the same tools the pros at doitong.com use, distilled into one always-available Chrome panel. ━━━ WHAT YOU CAN DO ━━━ 🎙️ VOICE MODE — Real Conversations, Real Tools Tap the mic, say "draw me a cosmonaut on a Tbilisi rooftop at golden hour," and watch it appear. Doitong Copilot speaks back in natural voice, asks clarifying questions, and fires off image and video generation jobs while you keep talking. Powered by Grok's realtime voice agent — under one second to first audio, full-duplex, no walkie-talkie awkwardness. 🖼️ IMAGE GENERATION & EDITING Generate images from text or edit ones you've already made. Just say "now make her hair red" or "add a cat next to him" and the copilot uses your last image as a reference. Backed by Nano Banana Pro (free tier) and ByteDance (premium quality). 🎬 VIDEO GENERATION Short cinematic clips from a prompt or from a still image. Choose duration, aspect ratio, and the model that fits your shot — Grok Video, Veo 3, Seedance 2, Kling, or Wan. Multi-reference support for showcase-style storyboards. 💬 CHAT WITH FOUR FLAGSHIP MODELS Switch between Claude (Anthropic), Gemini (Google), ChatGPT (OpenAI), and Grok (xAI) on the fly. One sign-in, one credit balance, no separate API keys to juggle. 🗂️ CONTEXT-AWARE HISTORY Open a past conversation and the copilot picks up where you left off — including voice mode, which seeds prior turns so you can verbally continue any thread. ⚡ ASYNC STUDIO STRIP Long-running generation doesn't block the chat. Pending jobs live as thumbnails at the top of the panel; finished media drops back into the conversation when ready, even if you've moved on to the next idea. 🔗 ONE-CLICK SHARING Every generation gets a public share URL on doitong.com. Copy, send, post — it just works. ━━━ WHO IT'S FOR ━━━ • Creators iterating on visual concepts who don't want to break flow. • Anyone who hates typing prompts and would rather just talk. • People who want Claude + Gemini + GPT + Grok in one place without managing four subscriptions. • Doitong.com users who want the full studio in their browser sidebar. ━━━ HOW IT WORKS ━━━ Sign in with your doitong.com account (email or Google). New accounts get free starter credits — enough to play with text chat, image generation, and a couple of video clips. Buy more inside the extension when you need them. Your prompts are sent to doitong.com's proxy, which forwards them to the underlying model and meters credits. We don't store API keys on your machine — your JWT is the only credential. ━━━ PRIVACY ━━━ • No data collection beyond what's needed to run a generation. • No analytics SDKs. • Voice audio is streamed to xAI's realtime endpoint and not retained by us. • Your generation history lives in your doitong.com account; delete it anytime. ━━━ PERMISSIONS ━━━ • microphone — only used in voice mode, only while the mic button is active. • host permission for doitong.com — required to call our LLM proxy and image/video endpoints. • storage — to keep your sign-in JWT and chat history local across sessions. ━━━ LINKS ━━━ • Web app: https://doitong.com • Privacy policy: https://doitong.com/policy • Support: support@doitong.com