GoBanana: Edit Photos with Text

You need to remove a stranger from a photo, swap a background or build a card from three shots, and you have not opened an editor in years. GoBanana does it with text: upload a photo, write what to change, get the result. Right in Telegram, with nothing to install. 🍌

Photoshop you talk to

The bot is an image editor built on neural networks. According to its profile, you can edit your photos with a text description, generate pictures from scratch and merge several photos into one.

Instead of layers and masks you simply state the task: "remove the sign", "make the lighting evening", "add a hat". The neural network decides how to do it.

Several neural networks under one roof

Inside you will find multiple models for generating and editing images. It helps: each has its own style, one holds a face better and another draws in a certain look. If a result disappoints, try a different model with the same request.

On top of that come 800-plus ready prompts (requests for the neural network). You do not have to invent wording, just pick a fitting one and fill in your details. 📋

Three trial attempts

The bot description says you get 3 free tokens at registration. That is enough to see if the service fits you. For later terms, open the profile inside the bot (/profile), where your personal account lives.

The first launch

1. Send /start and accept the terms of use.
2. Open /menu and pick a task: editing, generation or merging photos.
3. Upload a photo and describe what to change.
4. Judge the result and refine the request if needed.

If the bot keeps old context and gets confused, /reset clears it. Hints live in /guide and /help. 📖

Writing requests that work

• One edit at a time. Asking to change background, clothes and weather at once makes a mess.
• Be specific. "Warm sunset light from the left" beats "make it nice".
• Say what to keep. Name what must stay the same: face, pose, composition.
• Use a good source. A sharp photo gives a better result than a blurry one.

Ideas to start with

Try removing a random passer-by from a holiday shot, turning an ordinary selfie into a portrait in another style, or building a family collage from several pictures. The neural network handles such tasks fast, and they show what the service can do.

For product cards, a changed background and a highlighted item come in handy. Save your best phrasings in notes to repeat successful tricks. 💡

Remember that the result does not always land on the first try. Neural networks sometimes distort fingers, text in photos and small details, so check the picture before publishing.

Careful with other people's faces

When you edit photos of real people, ask for consent. Do not make fake pictures that may mislead or hurt someone. Neural edits are easy to pass off as real photos, and the responsibility is yours. ⚠️

Do not upload documents, pictures with personal data or other people's private shots: the image goes to a service for processing.

People who need to fix a personal photo fast, make a picture for a post or a product card, and prefer not to learn editors. For complex professional retouching with exact control, a regular graphics editor fits better.