MeiGen AI is an online AI photo editor and AI image generator that combines image-to-image editing and text-to-image generation in a single prompt bar. Users upload an image, write in natural language what they want to change, and then choose a suitable model, such as GPT Image 2.0, Nano Banana Pro, Seedream 5.0 Lite, or other models, to complete editing, generation, and export at up to 4K in the browser. If no photo is attached, the same prompt bar can also generate images from text alone. Credits cover all image and video models, with no need to manage API keys.
Main capabilities
- Image-to-image editing: After uploading an image, restyle a room, product, or portrait while keeping the layout and subject recognizable.
- Natural language photo editing: Describe a new sky, lighting, or season and apply the changes without brushes, masks, or layers.
- Text-to-image: Skip uploading and directly enter the subject, lens, and lighting, choose square, vertical, or wide compositions, and render at 1K, 2K, or 4K on supported models.
- Readable text within images: GPT Image 2.0 and Nano Banana Pro can lay headlines and labels directly into posters or packaging mockups.
- Multi-reference image blending: A single request can blend characters, products, locations, and style samples, with some models supporting up to 9 or 14 input images.
- Up to 4K resolution: Test ideas at 1K first, then render final results at 2K or 4K on supported models.
- Any aspect ratio: From square social posts and vertical stories to 21:9 ultra-wide banners, all can be set in the same menu.
How to use
- Create an account with Google or email, get free welcome credits, then scroll to the prompt bar on the homepage.
- Attach one photo to edit, attach multiple photos to combine, or attach no photo for direct text-to-image; the same prompt bar handles all three cases, and the model can be selected at the same time.
- Describe the changes in natural language: subject, environment, lighting, and style. You can include HEX colors or put text that must appear in quotation marks. Short and specific is better than long and vague.
- Compare results side by side, keep the best one, and adjust only one detail per follow-up, for example background first and then lighting.
- Save satisfactory images as PNG or WebP. All results are also saved along with the prompts in My Creations, making it easy to reopen and continue editing later.
Available models Image models include GPT Image 2, Nano Banana Pro, Nano Banana 2, Nano Banana, Seedream 5 Lite, Seedream 4.5, Grok Imagine, Wan 2.7 Image, and Qwen Image 2. Models are selected per request, allowing comparison of how the same prompt performs across different models. Video models include Veo 3.1 Fast, Veo 3.1 Quality, Sora 2, and Seedance 1.5 Pro, enabling short video generation with the same credit set.
Use cases
- Ad angle expansion: Starting from an approved key visual, request new scenes, times, camera angles, or art styles, suitable for paid social, display ads, and seasonal promotions.
- Product listing images and try-on concepts: Place in product images or model photos to generate clean listing images, lifestyle scenes, holiday backgrounds, or clothing try-on ideas.
- Story and game characters: Load multiple reference images (face, clothing, props, color palette) to keep the same character recognizable across different images.
- Everyday retouching: Describe edits in sentences, such as erasing passersby, warming a dull sky, expanding the canvas, or changing the color of a coat.
Limits and notes
- The maximum number of reference images varies by model: Nano Banana 2 and the Seedream series support up to 14, while Wan 2.7 supports up to 9.
- 4K resolution is supported only on some models; other models return their respective standard resolutions.
- Each generation consumes credits, with the amount depending on the selected model and settings.
- You can upload PNG, JPG, or WebP images.
- Commercial use is included in paid plans; responsibility for uploaded photos rests with the user.
- Uploaded content is used to perform the requested edits, and results are saved on the account's My Creations page; prompts and uploaded content undergo automatic content moderation.
- When using human faces, you should check similarity before publishing and upload only faces you have the right to use.
Plans and pricing Annual plans cost half the monthly price: Starter starts at $9.9 per month, billed annually. Starter includes 800 credits per month (9600 per year), roughly 80 images per month; Pro includes 2000 credits per month (24000 per year), roughly 200 images per month; Studio includes 6000 credits per month (72000 per year), roughly 600 images per month. With annual billing, the full year's credits are granted at once. All paid plans support commercial use and are watermark-free. The specific current plans are subject to the pricing page.
