Upload the original image
Start with a Gemini or Nano Banana image file instead of a blurry screenshot. Original files keep the lower-right mark, compression, and edge detail easier to inspect.
Upload a Gemini or Nano Banana image with the visible sparkle mark, preview a real before and after result, and download a cleaned copy without rewriting the whole picture. Use it for user-owned drafts, product mockups, and batch cleanup when you need to remove Gemini watermark marks before sharing.


The slider uses a live output from the remover, with the visible Gemini mark cleaned from the lower-right corner.
The generator is separated from the hero so users can inspect the examples first, then upload one image or a small batch when they are ready.
One file uses the single-image API. Two or more files use the batch endpoint with server-side concurrency control.
Each image costs 2 credits. Files that fail because no visible Gemini mark is detected are refunded by the backend.
Successful and failed attempts are stored as task and image records for admin history, usage stats, and future user history views.
Drop one image for a single cleanup or upload several files for batch removal. The tool charges 2 credits per image and refunds failed items when no visible Gemini watermark is detected.
Drag images here or choose files from your device. Use original downloads when possible because screenshots, heavy compression, or cropped corners can make detection less reliable.
PNG, JPG, JPEG, WebP, or BMP. Up to 10 images, 20MB each.
Upload a Gemini image with a visible sparkle watermark. The cleaned output will appear here after processing.
Single image and batch removal use the same backend queue. Batch files are processed server-side with a small concurrency limit for VPS stability.
Drag each slider to compare the original upload against the cleaned output. These examples show product scenes, portraits, and layout-style images handled by the same remover API.


A clean corner keeps the product scene usable without cropping away the flowers or shadow.


The remover targets the visible sparkle mark while preserving the product and backdrop.


Useful for creator drafts where the lower-right mark distracts from the full composition.
The workflow is intentionally simple: upload the image, detect the visible mark, then download a cleaned copy. It is not a prompt-based image editor, so there is no need to write instructions like remove the watermark.
Start with a Gemini or Nano Banana image file instead of a blurry screenshot. Original files keep the lower-right mark, compression, and edge detail easier to inspect.
The remover looks for the Gemini sparkle watermark area. If no visible mark is found, that file is marked as failed and the image credit is refunded.
Successful files return a cleaned image URL and are also saved into the existing image records, so admin history and usage statistics can inspect the task.
For multiple images, upload them together and let the server process the batch. The page does not fire a browser Promise.all burst because small VPS deployments need controlled concurrency.
Two or more files use the batch API. The backend creates one batch task, one image record per file, and a clear success or failed state for each item.
The batch cost is 2 credits per image. Failed files or files with no visible Gemini mark are refunded automatically by the backend.
Because the results are stored as tasks and images, batch usage can later power admin search, usage reporting, and a My Works view if you decide to expose it.
Use the Gemini watermark remover when the image is yours to edit and the visible mark is the problem. It works best as a cleanup step after image generation, not as a universal logo eraser.
Clean Gemini image drafts before sending them to a client, posting a concept, or placing them into a social layout where the corner mark distracts from the design.
Remove Gemini watermark marks from AI product scene drafts without cropping away important bottom-corner composition, callouts, shadows, or packaging detail.
Process several candidate images together, then keep only the outputs that still look clean after the visible mark is gone.
The current Gemini watermark remover is deliberately scoped. It should make the visible Gemini mark less disruptive, but it should not be treated as a rights, attribution, or authenticity bypass.
This page supports image upload and does not remove video watermarks, audio metadata, or frame-by-frame marks from Gemini, Veo, Omni, or other video tools.
It does not remove SynthID, invisible watermarks, provenance metadata, or arbitrary third-party logos. It also will not clean every fake or hallucinated watermark perfectly.
Check the lower-right corner before publishing. If the source image is very compressed, heavily cropped, or contains important detail under the mark, the result may need manual review.
You can sign in and use your account credits. The current cost is 2 credits per image, and failed files are refunded when the tool does not detect a visible Gemini watermark.
Yes. Upload 2 to 10 images and the batch endpoint will process them with server-side concurrency control. This is better for a 2C4G VPS than launching many browser requests at once.
No. This page is for the visible Gemini sparkle mark on image files. It does not remove SynthID, invisible provenance signals, image metadata, or platform-level detection.
The image may not contain the supported visible mark, the mark may be cropped out, or compression may have changed the corner too much. In that case the item is marked failed and its image credit is refunded.
The tool is designed to target the visible mark area instead of re-generating the full image. You should still compare the output with the original before publishing.
No. Use it only on images you own or have permission to edit, and only for the supported visible Gemini mark. It is not a general-purpose copyright or brand watermark remover.