SuperAPI AI Directory

anyimg.ai

anyimg.ai is a prompt-based AI image editor that completes image editing through natural language instructions. Core capabilities include: removing objects from the frame, replacing backgrounds, retouching portraits, restyling, and generating new images. The tool supports reading and preserving model pose and lighting, extracting clothing graphics and colors, checking the source consistency of generated results, and can modify only selected areas during editing while protecting composition, identity features, lighting, and unselected pixels unchanged. It supports uploading reference images of people, products, or brands, allowing the editor to extract key cues and output a stylistically unified editable image set. The editor provides a Version Trail feature, where every modification becomes a visible branch, allowing comparison of different directions, returning to the original version, or continuing to move forward. It supports models such as GPT Image 2, Nano Banana Pro, and Nano Banana 2, which can be switched per use or automatically selected by the agent. Registration comes with free credits, and paid plans start at $4.90 per month, including commercial licensing and unlimited downloads. Paid plans are billed by credits, with credit consumption varying by model and resolution. Credits are automatically refunded for failed tasks, and subscriptions can be canceled at any time in billing settings.

Open website

About this tool

anyimg.ai is a conversational AI image editor (AI Image Editor) focused on prompt-based image editing. Users only need to describe the desired modification in natural language, and the agent will automatically call the most suitable model to execute it.

[Purpose] The tool is aimed at users who need to quickly complete commercial-grade image editing and generation, covering directions from portrait retouching to product visuals, brand campaigns, covers, spatial scenes, and more. The official emphasis is that its examples are art-directed with usable commercial assets as the goal, rather than just style demonstrations.

[Main capabilities]

  • Remove objects, replace backgrounds, retouch portraits, restyle photos, and generate new images.
  • Read model, pose, and lighting; extract T-shirt prints, colors, and cuts; generate and check source consistency.
  • Can combine multiple reference images for compositing. Example: use the model from image 1 and the amber perfume from image 2 to create a premium perfume advertisement, preserving the person's identity as well as the bottle's exact shape, cap, and color.
  • Point-based regional editing: select an area and describe it in natural language, and the editor will modify only the target area while protecting composition, identity, lighting, and untouched pixels. Example: paint only the lip area and apply vivid red lipstick, keeping identity, pose, clothing, background, and lighting unchanged.
  • A direction approved once can be extended into multiple deliverables: key advertising visual, social assets, website header image, product detail image.
  • Version Trail: every edit becomes a visible branch, allowing comparison of directions, returning to the original version, or continuing to move forward; unlike a flat undo stack, it keeps the reason, source, and result of each branch together.

[How to use] Users describe editing needs in natural language and upload reference images of people, products, or brands. The agent will preserve key cues and output unified, editable results. Users can switch models for each edit, or let the agent choose automatically.

[Supported models] Currently supports GPT Image 2, Nano Banana Pro, and Nano Banana 2, with more models to be added over time.

[Pricing and credits]

  • Free credits are given upon registration, and all supported models can be used.
  • Paid plans start at $4.90 per month, and every paid plan includes commercial use authorization and unrestricted downloads.
  • Credits are consumed based on model and resolution; high resolution and advanced models may consume more credits; if a generation task fails, the consumed credits are automatically refunded.
  • Examples of plan credits officially listed: 1,300 credits (valid for 1 month, up to about 433 standard edits), 3,090 credits (up to about 1,030 times), 5,780 credits (up to about 1,926 times), 12,060 credits (up to about 4,020 times), with standard edits calculated at 3 credits per use.
  • You can cancel your subscription at any time in the Billing settings, and you can continue using it until the end of the current billing cycle.

[Use cases]

  • Product photography and publishing assets: such as specialty coffee packaging and displays, sneaker product visuals, and perfume studio key visuals.
  • Brand and advertising creative: such as advertising campaign concepts, jewelry advertising portraits, and performance ad art direction.
  • Portraits and creative visuals: such as creative portraits, high-end pet care advertising portraits, and architectural light-and-shadow fashion editorial images.
  • Space and scene generation: such as minimalist home lifestyle scenes, coastal gallery interior concepts, and tech website hero image backgrounds.
  • Packaging and displays: such as Oriental tea packaging displays.

[Limitations and notes]

  • The official source does not provide specific details such as the exact quota amounts for non-paying users, generation speed limits, supported image formats, or maximum resolution.
  • The specific values for credit consumption are not fully disclosed; it only states that they depend on the model and resolution, with advanced models and high resolution consuming more.
  • The official source does not provide a specific model list beyond GPT Image 2, Nano Banana Pro, and Nano Banana 2.
  • The official source does not state whether features such as an API or team collaboration are provided.
  • After canceling your subscription, you can still access it until the end of the current billing cycle; details such as whether generated content is retained after the cycle ends are not provided.