ReelMintreelmint

AI Image Generator and Editor: Text to Image, Image to Image, and What Each Model Costs

Text to image and image to image on Nano Banana, Seedream, GPT Image, FLUX, Qwen, Ideogram and more, from 1 credit an image. Steps, limits, prices.

To generate an image with AI in ReelMint, open Mint Studio, pick Text to Image, write what you want to see, choose a model and a size, and generate; to edit a picture, pick Image to Image instead, add one or more source images and describe the change. Prices run from 1 credit an image on Z-Image to 30 credits for a 4K image on Nano Banana Pro, and every price is on the button before you click.

We build ReelMint, an AI studio that makes videos, images and music and posts them to TikTok, YouTube, Instagram, Facebook and X. There are 28 text-to-image models and 26 image-to-image models in the studio today, and every plan runs every one. This post lists them with their credit prices and says which to pick for what.

Text to image, step by step

  1. Open Mint Studio and choose Text to Image.
  2. Write the picture. Subject, setting, light, style, framing. Be literal: the model draws what is written.
  3. Pick a model and a size. Many models price by output size (1K, 2K, 4K); the button updates as you choose.
  4. Generate. The image lands in your Library, where it can be downloaded, posted, or used as the start frame of a video.

The text to image tutorial covers picking a model, prompting for a style and saving images for later videos.

Text-to-image models and prices

  • Z-Image, the cheapest image: 1 credit
  • FLUX 3 Klein, the cheapest 4K: 2 credits at 1K, 4 at 2K, 9 at 4K
  • Nano Banana 2.1: 5 credits at 1K, 8 at 2K, 12 at 4K
  • Nano Banana 2 Lite: 5 credits at 1K
  • Grok Imagine: 5 credits an image, 6 takes per run; Grok Imagine Image 2.0 for sharper text, 5 credits
  • Seedream 5.0 Flash: 5 credits up to 2K; Seedream 5.0 Lite: 7 credits up to 4K; Seedream 5.0 Pro: 9 at 1K, 18 at 2K; Seedream 4.5: 9
  • Qwen Image: 5; Qwen Image 2.1: 5 at 1K, 10 at 2K; Qwen Image 3.0: 6; Qwen Image 3.0 Pro: 8 at 1K, 15 at 2K
  • Imagen 4 Fast: 5; Imagen 4: 10; Imagen 4 Ultra: 15
  • Ideogram V3, the best text rendering: 5 turbo, 9 balanced, 13 quality
  • GPT Image 2, GPT Image 2.5 Flare and GPT Image 2.5 Sunburst: 8 at 1K, 13 at 2K, 20 at 4K; GPT Image 1.5: 5 medium, 28 high
  • FLUX.2 Pro: 7 at 1K, 9 at 2K; FLUX.2 Flex: 18 at 1K, 30 at 2K
  • Wan 2.7 Image: 6; Wan 2.7 Image Pro: 15
  • Nano Banana Pro: 23 at 1-2K, 30 at 4K

Prices are per image as the studio shows them; the live table is on the models page.

Which image model to pick

  • Many drafts, lowest cost: Z-Image or FLUX 3 Klein.
  • Words in the picture (a poster, a sign, a thumbnail title): Ideogram V3 or Grok Imagine Image 2.0.
  • A 4K still for print or a large screen: FLUX 3 Klein at 9 credits is the cheapest 4K; Nano Banana 2.1 and the GPT Image models also go to 4K.
  • Several options from one prompt: Grok Imagine returns 6 takes per run.

AI image editing: image to image

Image to Image changes, restyles or extends a picture you already have. Upload a picture or choose one from your Library, describe the change ("same person, same pose, now in a rainy street at night"), and say what must stay the same. The image editing tutorial shows it step by step.

The editing models and prices:

  • Several source images in one edit: Seedream 4.5 Edit and Seedream 5.0 Lite Edit take up to 14 images (9 and 7 credits); Nano Banana 2.1 Edit, FLUX 3 Klein Edit, Seedream 5.0 Flash Edit, Seedream 5.0 Pro Edit and Qwen Image 2.1 Edit take up to 10; Qwen Image 3.0 Edit and 3.0 Pro Edit take up to 3, plus 1 credit per input picture.
  • General edits: Grok Imagine Edit and Grok Imagine Image 2.0 Edit (5), Qwen Image-to-Image (5), Qwen 2 Image Edit (8), Qwen Image Edit (8 per megapixel), FLUX.2 Pro Edit (7 at 1K), GPT Image 2 Edit and the GPT Image 2.5 edits (8 at 1K), Nano Banana Pro Edit (23 at 1-2K).
  • Consistent faces: Ideogram Character, 15 turbo, 23 balanced, 30 quality.
  • Remix a picture's style: Ideogram V3 Remix, 5 to 13.
  • Utilities: Recraft Remove Background (2 credits), Recraft Crisp Upscale (1 credit), and the Topaz Image Upscaler (13 credits up to 2K, 25 at 4K, 50 at 8K output).

From still to video

A generated image is also the cheapest way to control a video. Make the still until the face, outfit and place are right, then animate it with image to video: the photo locks the look and the prompt sets the motion. See image to video.

Is there a free AI image I can claim?

Yes. The free examples include finished images with their prompt and model, and claiming them needs no card. Your own prompts run on a plan: Lite is $9.99 a month for 800 credits, which is hundreds of images on the 1 to 5 credit models.

Are the models restricted by plan?

No. Every plan runs every model, from Z-Image to Nano Banana Pro. The only difference between plans is how many credits you get.

What happens if an image fails?

A provider rejection is refunded in full. You are charged only for what was actually produced.

Can I write the prompt with AI?

Yes. AI Prompt sharpens a rough idea into a detailed image prompt for a few credits, shown on the button.

Keep reading

Make your own

Claim a free example with no card, or start on Lite: $9.99 a month, 800 credits, every AI model.

Get started