Free AI image editing prompts. Compatible with our Quickstart workflows. Copy, tweak, and use with either the cloud client or supported offline alternatives.
The seed, steps, guidance, and strength examples below are starting points for compatible open-model pipelines. They are not Gemini API parameters.
replace background with neon cyberpunk city, preserve subject lighting and pose
Open-model starting point: seed=42 steps=30 guidance=4.0 strength=0.6
convert outfit to modern tactical style, keep face identity and hair
Open-model starting point: seed=42 steps=30 guidance=4.0 strength=0.55
add a katana in right hand, realistic reflection and shadow
Open-model starting point: seed=64 steps=32 guidance=4.5 strength=0.5
keep character identity and facial features; change background to sunny beach, match lighting and shadows
Open-model starting point: seed=21 steps=28 guidance=4.0 strength=0.55
in masked area: refine hands and fingers with realistic skin texture and natural pose
Open-model starting point: seed=42 steps=24 guidance=4.2 strength=0.45
swap jacket to red leather while preserving scene composition, camera angle, and lighting
Open-model starting point: seed=77 steps=30 guidance=4.0 strength=0.5
apply subtle cinematic color grading; keep texture and identity unchanged
Open-model starting point: seed=9 steps=22 guidance=3.8 strength=0.35
place product on white seamless background with soft natural shadow; preserve geometry
Open-model starting point: seed=56 steps=26 guidance=4.2 strength=0.5
These prompts target documented model capabilities. HunyuanImage 3.0 Instruct supports multi-image fusion and reasoning-assisted editing; Qwen-Image-Edit-2511 focuses on consistent image editing.
HUNYUANIMAGE 3.0 INSTRUCT
Apply the style of Image 2 to the subject in Image 1. Preserve the pose and facial features from Image 1 while adopting the color grading and atmosphere from Image 2.
Uses 2 input images · CoT reasoning enabled
HUNYUANIMAGE 3.0 INSTRUCT
Analyze the lighting direction in the original image, then add a glass sphere on the table that reflects the same lighting conditions and casts a physically accurate shadow.
Model analyzes first, then executes · No mask needed
QWEN-IMAGE-EDIT-2511
Add a neon sign on the building wall that reads "CAFE OPEN 24H" in Art Deco style typography. Match the perspective and lighting of the scene.
Strong text generation capability · Precise typography
QWEN-IMAGE-EDIT-2511 / GLM-IMAGE
Convert this magazine layout into a scientific infographic about climate change. Include charts, data visualizations, and maintain the grid structure with proper alignment.
Structured output · Document understanding
text‑guided background replacement · local edits · identity/character consistency · mask‑guided inpainting · scene preservation · subtle style change · product photo shadows · multi-image fusion · CoT reasoning · complex text rendering