JiMeng (即梦) is ByteDance’s powerful AI video generation app — and its Camera Mode lets you upload a photo or short clip, then generates a polished, talking-head-style video featuring *your* face or likeness. It’s incredibly useful for creators, educators, and marketers building AI-powered content fast.
But there’s one big catch: every video exported from JiMeng’s Camera Mode carries a prominent, semi-transparent watermark — usually placed in the bottom-right corner. Worse, it’s baked directly into the video frame. That means standard “crop-and-scale” tricks won’t work cleanly. And most generic watermark removers? They blur, distort, or leave ghosting artifacts — especially around facial details or motion areas.
If you’ve tried editing out the JiMeng watermark manually — or used AI inpainting tools that degrade resolution — you know how frustrating it is to lose sharpness, smooth skin tones, or crisp text overlays.
The good news? You *don’t* need to sacrifice quality. There’s a direct, lossless way to get the original, unwatermarked source file — 100% native resolution, zero re-encoding, no blurring, no artifacts.
Here’s exactly how.
---
Before jumping into the solution, it helps to understand *why* most approaches fall short.
JiMeng (即梦) doesn’t just overlay a simple logo. In Camera Mode, the watermark is embedded at the CDN or API level — meaning it’s part of the final encoded stream delivered to your device. When you download or share a video from JiMeng, what you get is *already watermarked*. So:
What you really need isn’t *removal* — it’s *extraction*: pulling the clean, pre-watermarked version straight from JiMeng’s infrastructure.
That’s where OffWatermark comes in.
Unlike desktop editors or browser-based editors that process your uploaded file, OffWatermark works server-side. It takes your JiMeng share link — like `https://jimeng.jianying.com/s/abc123` — and uses reverse-engineered, platform-specific extraction logic to fetch the original, unwatermarked master asset *before* the watermark layer was applied.
No pixel manipulation. No generative fill. Just the raw source — untouched and pristine.
And yes: this applies *only* to videos generated in JiMeng’s Camera Mode, since that’s where the watermark appears. Other modes (e.g., text-to-video or image-to-video without face input) may not include it — but if you see it, OffWatermark handles it.
*(Note: Dreamina is the international version of JiMeng — same underlying tech, same watermark behavior. OffWatermark supports both.)*
---
This isn’t an app. No download. No install. No permissions. Just your phone, a browser, and 50 seconds.
> 💡 Pro tip: If you’re using Dreamina (the global JiMeng version), the share link format is identical — same extraction logic applies.
That’s it. No waiting for AI to “process.” No progress bar spinning for minutes.
OffWatermark instantly routes your request to its server infrastructure, authenticates the link against JiMeng’s public API endpoints, and retrieves the original MP4 source — the exact file JiMeng used *before* applying the watermark layer.
No re-encoding means:
You can drop this file straight into CapCut, Premiere, DaVinci Resolve — or upload it watermark-free to TikTok, YouTube, or your portfolio site.
---
While JiMeng (即梦) Camera Mode is a standout use case — especially with rising demand for AI avatar videos — OffWatermark wasn’t built for just one app.
It supports five major platforms, all with the same zero-compromise workflow:
| Platform | Watermark Type | Example Link |
|----------|----------------|--------------|
| JiMeng (即梦) | Camera Mode AI watermark (bottom-right, semi-transparent) | `https://jimeng.jianying.com/s/xxxxx` |
| Dreamina | Identical to JiMeng — same removal logic | Same format as above |
| Douyin (抖音) | Platform logo + user ID (often bottom-left) | `https://v.douyin.com/xxxxx` |
| TikTok | “tiktok.com” logo + creator handle (top-right or bottom) | `https://www.tiktok.com/@user/video/xxxxx` |
| Kuaishou (快手) | “Kwai” logo + channel ID | `https://v.kuaishou.com/xxxxx` |
| Xiaohongshu (小红书) | “RED” or “XHS” watermark + username | `https://www.xiaohongshu.com/explore/xxxxx` |
Why does this matter? Because creators often cross-post — generate in JiMeng, share to Douyin *and* TikTok, repurpose clips for Xiaohongshu carousels. With OffWatermark, you use *one workflow*, one login, and the same high-fidelity output — whether it’s a 10-second AI talking head or a 3-minute lifestyle tutorial.
No switching tools. No learning new UIs. No quality trade-offs.
Just paste → extract → save.
Pricing stays simple too:
All plans give you the same technical result: 100% original quality, zero artifacts, zero re-encoding.
---
> Disclaimer: OffWatermark is an independent tool and is not affiliated with, endorsed by, or connected to ByteDance, JiMeng (即梦), Dreamina, Kuaishou (快手), or Xiaohongshu (小红书) in any way. All trademarks belong to their respective owners. Users are solely responsible for ensuring their use complies with applicable laws and terms of service. Only remove watermarks from videos you personally created.
← Back to OffWatermark