AI Images for TikTok, Made in Telegram

Make vertical AI images for TikTok without leaving a Telegram chat. Describe a slide, a cover or a scene, or attach a selfie and name the trend, and the bot sends back a 9:16 image in about a minute, sized for a phone screen. Your first 5 credits are free, and there is no app to install.

A TikTok AI image generator that lives in a chat

TikTok is a vertical feed, and most image generators default to square. Here you pick 9:16 once in the model menu and every result fills the screen. The rest is a normal conversation: you write what you want, the picture arrives, and if something is off you reply with the correction.

The images people make for TikTok fall into four groups:

• Photo mode slideshows and carousels. A set of vertical slides with a shared look, for a story, a list, a mood board or a “day in the life”.
• Video covers. A strong first frame or thumbnail, with or without a title on it.
• AI photo trend edits of yourself. Your own selfie turned into a film shoot, a new outfit, a new place or a new era.
• Backgrounds for text on screen videos. A calm frame with empty space where your caption or voiceover text will sit.

The examples above cover the last three, and the exact prompts that made them are below.

The bot is an independent Telegram bot and is not affiliated with, endorsed by or connected to TikTok or ByteDance. The name is used here only to describe where the images end up.

How to make AI images for TikTok in Telegram

1. Open the bot in Telegram and press Start.
2. Claim 5 free credits. Subscribe to @zosee_channel, then return to the bot and tap the claim button. The subscription alone grants nothing, the tap is what releases the credits. See the free credits page.
3. Pick a model and the 9:16 aspect ratio in the bot’s menu.
4. Send your prompt. For a trend edit, attach your selfie and write the instruction as the caption of that same message. A photo alone, or a prompt sent as a second message, will not be paired.

Up to 4 photos fit in one message, which helps when one selfie has your best expression and another has better light. Save the results to your phone and they are ready for the TikTok upload screen.

The TikTok AI photo trend, from your own selfie

Most AI photo trends on TikTok follow the same shape: an ordinary photo of you, then the same face somewhere far more interesting. That is an image edit, and it needs two things in the prompt: the new scene, and one line that protects your face. This is the prompt behind the desert road example, unchanged:

Turn the attached selfie into a 1990s film photo shoot: the same woman leaning on an old unbranded car with no badges or logos on an empty desert road at golden hour, wind in her hair, wearing the same denim shirt. Warm faded film colours, soft grain, slight light leak at the edge, vertical 9:16 frame. Keep her face, freckles and hair exactly as in the attached photo.

Swap the scene for your own idea and keep the last sentence. “No badges or logos” is there because a vintage car otherwise tends to come back with a real maker’s badge on the grille. The starting selfie in the examples was itself generated with this prompt, so you can see the before and after as a pair:

Casual vertical phone selfie of a young woman in her twenties with shoulder-length wavy auburn hair and light freckles, wearing an oversized light blue denim shirt, standing in a small apartment hallway with a coat rack behind her. Flat indoor ceiling light, slightly soft focus, natural skin, relaxed half smile, arm-length framing from the chest up. Ordinary unedited smartphone snapshot.

The result is a generated image, not a photograph of you. Label AI content where TikTok asks you to, and do not use someone else’s face without their consent.

AI slideshows and backgrounds for TikTok

A slideshow is a set, so the trick is consistency. Write the subject first and the style last, then keep the style half identical across every slide and change only the subject. This is the prompt behind the rainy window frame:

Vertical aesthetic slideshow frame: a cozy reading corner by a rainy window at night, a steaming mug of tea on a wooden sill, an open book, warm string lights reflected in the wet glass, a sleeping cat on a knitted blanket. Soft film look, warm amber and deep blue palette, gentle grain. Leave calm empty space in the upper third for text on screen. No text, no letters, no logos.

Two parts of it matter for TikTok specifically. “Leave calm empty space in the upper third” gives your on screen text somewhere readable to sit, and “no text, no letters” stops the model from inventing lettering you would have to cover up. Remember that TikTok’s own buttons and caption cover the right edge and the bottom of the screen, so keep faces and key objects near the centre. For more directions, the wallpapers category in the prompts library is already vertical and works well as slideshow material.

TikTok cover images

A cover has one job: to make someone stop scrolling. Ask for one clear subject, big and centred, with strong contrast against the background. If you want the title inside the image, use GPT-Image-2 and write the exact words in quotes, for example a bold title that reads "3 cheap dinners". If you prefer to add the title in TikTok’s editor, ask for empty space at the top and no text in the image.

Which model to pick for TikTok images

• NanoBanana 2 is the everyday pick and made the selfie trend example: 2 credits at 1K, 3 at 2K, with the best balance of quality and price for edits of your own photo.
• Seedream is the stylized one and made the slideshow frame. It follows loose, moody prompts well and runs from 2K to 4K at 2 to 4 credits.
• NanoBanana Pro is the one for a crowded scene or a demanding edit where the face has to hold perfectly, 4 to 5 credits.
• GPT-Image-2 is the cheapest at 1 credit for 1K and the best at legible text, so it is the natural choice for a cover with a title on it.

All four, and Qwen3, offer the vertical 9:16 ratio.

What it costs

1 to 6 credits per image depending on the model and resolution, and credits never expire. A ten slide carousel at 1K on GPT-Image-2 is 10 credits, the same set on 2K Seedream is 20. Payment is Telegram Stars, cryptocurrency, a bank card from any country ($), or a Russian bank card (₽). The pricing page has the packs.

Turn your TikTok images into video

TikTok is a video platform, and a still frame is often only the start. The rain on the window can run, the steam can rise from the mug, the wind can move the hair in the desert shot. Our sister video bot does that part, also inside Telegram. Save the image you made here, attach it in the video bot, and write the motion as the caption of the same message. The photo costs nothing extra there, only the seconds of video are billed.

Describe the motion rather than the picture, since the model can already see the frame (“slow push in, rain running down the glass, steam rising from the mug”, not “a cat by a window”), and start with a short clip so the result stays close to your image. The walkthrough is on the image to video in Telegram page.

Related

• AI avatar generator, for a profile picture made from your own selfie.
• AI photo editor, for edits to a photo you already have, not a whole new scene.
• Change or remove a photo background, to put yourself somewhere else in one step.
• Photo to anime, for the anime version of the selfie trend.
• How to use the bot, the walkthrough from Start to your first image.
• Instagram photo generator, for aesthetic feed posts and Stories from one selfie.
• YouTube thumbnail generator, for 16:9 thumbnails with bold readable text.
• Ghibli style generator, for soft hand-painted Ghibli style art from a photo or text.