AI Image Editor
Edit · Preserve · Combine ReferencesUpload a photo or several references, describe the exact change, and preserve identity, products, typography, layout, lighting, or composition with an editing-capable AI model.
AI image editing examples
Separate the requested transformation from the identity, geometry, labels, layout, and lighting that must remain fixed.
Assign each image a role such as base scene, product, pose, palette, texture, style, or layout.
Create and revise posters, labels, UI, diagrams, packaging, and localized graphics with exact quoted copy.
The picker and catalog include only live image models whose input contract accepts images.
A controlled image editing workflow
- 1Upload the source and references
Use the plus button to attach the base photo and any product, style, pose, palette, texture, or layout references.
- 2Write change and preserve rules
State exactly what changes, then inventory the face, product, text, geometry, camera, lighting, and surrounding details to preserve.
- 3Choose an editing model and refine
Compare only image-input models, generate the edit, and make one precise correction at a time.
Choose an AI image editing model
Black Forest LabsFLUX.2 [klein] 9B- Modalities
- text / image → image
- Released
- Jan 23, 2026
- Capabilities
- Text To Image · Image To Image · Edit
- References
- Up to 4 images
- Resolution
- 128–2048 px
- Formats
- JPG · PNG · WEBP
Ultra-fast image generation and editing with sub-second latency
Black Forest LabsFLUX.1 Schnell- Modalities
- text → image
- Released
- Aug 1, 2024
- Capabilities
- Text To Image
- Resolution
- 128–2048 px
- Formats
- JPG · PNG · WEBP
Ultra fast FLUX.1 text to image model for local use
Stability AISDXL- Modalities
- text / image → image
- Capabilities
- LoRA · ControlNet · Inpainting · Styles
Flexible SDXL workflow with checkpoint, LoRA, ControlNet, inpainting, and style controls.
Stability AIStable Diffusion 1.5- Modalities
- text / image → image
- Capabilities
- Checkpoints · LoRA · Inpainting
Classic SD 1.5 workflow for fast image generation with many checkpoints and LoRAs.
OpenAIGPT Image 2- Modalities
- text / image → image
- Released
- Apr 21, 2026
- Capabilities
- Text To Image · Image To Image · Edit
- References
- Up to 16 images
- Resolution
- 16–3840 px
- Formats
- JPG · PNG · WEBP
Image generation and editing model with strong prompt fidelity, text rendering, and layout-aware control
OpenAIGPT Image 1 MiniSmaller OpenAI image model for faster, cheaper image generation and reference-based edits.
AlibabaQwen Image Edit 2511Qwen edit model for prompt-based image changes while preserving source structure and details.
PramaRemove background (BEN2)Background removal model that isolates the subject and outputs a clean cutout from an image.
TencentRestoring old photo- Modalities
- image → image
- Capabilities
- Face restoration · Upscale
- Input
- Image required
Face restoration model for repairing old, blurry, damaged, or low-resolution portraits.
Jason AnticColourising old photo- Modalities
- image → image
- Capabilities
- Photo colorization · Restoration
- Input
- Image required
Photo colorization model for adding natural color to black-and-white or faded images.
GoogleNano Banana 2- Modalities
- text / image / video → image
- Released
- Feb 26, 2026
- Capabilities
- Text To Image · Image To Image · Edit
- References
- Up to 14 images
- Resolution
- 0.5K · 1K · 2K · 4K
- Formats
- JPG · PNG · WEBP
Gemini 3.1 Flash Image fast high quality AI image generation and editing
ByteDanceSeedream 5.0 Lite- Modalities
- text / image → image
- Released
- Feb 24, 2026
- Capabilities
- Text To Image · Image To Image
- References
- Up to 14 images
- Formats
- JPG · PNG · WEBP
Responsive text-to-image generation with real-time search and precise prompt adherence
AlibabaQwen Image 2.0- Modalities
- text / image → image
- Released
- Feb 10, 2026
- Capabilities
- Text To Image · Image To Image · Edit
- References
- Up to 3 images
- Resolution
- 1–4096 px
- Formats
- JPG · PNG · WEBP
Unified image generation and editing with professional text rendering
xAIGrok Imagine Image- Modalities
- text / image → image
- Released
- Jan 29, 2026
- Capabilities
- Text To Image · Image To Image
- References
- Up to 3 images
- Resolution
- 1K
- Formats
- JPG · PNG · WEBP
AI image generation from text and images
Black Forest LabsFLUX.2 [pro]- Modalities
- text / image → image
- Released
- Nov 26, 2025
- Capabilities
- Text To Image · Image To Image · Edit
- References
- Up to 9 images
- Resolution
- 256–2048 px
- Formats
- JPG · PNG · WEBP
High control FLUX.2 Pro image generation and editing
Black Forest LabsFLUX.2 [flex]- Modalities
- text / image → image
- Released
- Nov 25, 2025
- Capabilities
- Text To Image · Image To Image · Edit
- References
- Up to 10 images
- Resolution
- 256–2048 px
- Formats
- JPG · PNG · WEBP
Configurable FLUX.2 Flex for precise text aligned images
GoogleNano Banana Pro- Modalities
- text / image → image
- Released
- Nov 20, 2025
- Capabilities
- Text To Image · Image To Image · Edit
- References
- Up to 14 images
- Resolution
- 1K · 2K · 4K
- Formats
- JPG · PNG · WEBP
Nano Banana Pro image preview for precise visual control
Showcases
Explore the same live showcase collection used on the GizAI home page.
AI Action Figure Generator
Turn yourself into an action figure. The showcase opens with FLUX.2 Klein for fast reference-image editing, and subscribers can switch to GPT Image 2 or Nano Banana 2 for higher-fidelity results.
InputAI Action Figure Video
Transform your action figure image into a video.
Consistent Character Story
Develop stories with visually consistent characters throughout. Upload your own character or generate one with AI.
AI Ghibli Style
Transform your photo into a Ghibli style image.
Create Complete Presentations
Generate professional presentations with customizable slides, graphics, and content. Export as PPTX for further editing.
Generate Professional Diagrams
Create flowcharts, UML diagrams, network layouts, and other visual representations from text descriptions.
Create Visual Thought Maps
Generate interactive mindmaps to organize concepts, brainstorm ideas, or visualize complex topics.
AI image editor FAQ
Can I edit an image without signing up?
Some editing-capable models include anonymous use. Current access and request limits appear on each model card before you continue.
How do I keep a face, product, or layout unchanged?
Name the source image as the base, describe only the requested change, and explicitly list identity, geometry, labels, pose, camera angle, layout, lighting, and surrounding objects to preserve.
Can I use multiple reference images?
Yes, when the selected model supports multiple images. Give every upload a clear role such as base scene, product, garment, pose, palette, style, texture, or layout.
Which models appear in the image editor?
Only live image-generation models whose actual field contract accepts image input. Text-only image models are excluded from both the picker and catalog.
Can AI edit text inside an image?
Models suited to visual text can revise posters, labels, packaging, UI, and localized assets. Quote exact copy, define placement and hierarchy, and forbid extra words.
What makes a reliable edit prompt?
Use a change-and-preserve contract: change only the named element, preserve the listed invariants, match physical lighting and shadows, and make one correction per revision.
What image sizes and formats are available?
Size, quality, output format, background, reference count, and variation count come from the selected editing model and appear in its live settings.