Conversational image editing
Describe the photo edit you actually need
Plan a photo edit in everyday language. Define the intended change, identify what should be preserved and compare the edited result with the original.
Model
This creator generates a still image. Continue in the linked tools for comics and video; research strategies are not selectable production modes.
Image record
Google Vertex AI · gemini-3.1-flash-image · · 1280 × 859 px
Original generation prompt
Use case: photorealistic-natural / product-mockup. Create a natural window-lit editorial PHOTOGRAPH of exactly TWO anatomically normal adult hands holding one landscape matte comic proof sheet above a real warm oak desk. Image1 supplied is the original adult botanist and mint robot reference sheet; image2 is the COMPLETE original SIX-PANEL greenhouse comic. The held paper must have landscape 3:2 proportions so Image2 fits WITHOUT rearrangement. Put the exact complete Image2 sequence on the held sheet as THREE columns by TWO rows, with a clean ivory outer margin: upper row arrival, pot on bench, gloved seed close-up; lower row watering, seedling reaction, final greenhouse view. Preserve all six distinct scenes in that order; DO NOT repeat watering, remove seed, reformat into two columns, add text or combine panels. Keep the original adult's teal jacket, coral scarf, dark curly hair, sand trousers and original mint round robot. Hold the paper by its two lower blank outside corners only; all artwork corners remain visible and hands cover no panels. The source character-reference sheet from Image1 rests flat farther back on the desk, partly visible beside the held print. Only the printed artwork is ink/watercolor illustration; hands, table, thin paper grain and room are a convincing photograph. No visible faces, booklet, screen, logos, titles, numbers, writing, fake UI, metrics or product packaging. Natural skin/five fingers, coherent optical perspective, subtle paper flex, quiet soft daylight. The completed hand-held story takes 75 percent of frame. Fictional lifestyle presentation only, never a customer photograph or research result. Landscape 3:2.
View full artworkPlan a specific result
A different jacket, the same original character
Plan an illustrative edit that changes an original adult character’s jacket color while retaining their face, pose and scarf. Use this as a controlled task with a clear comparison.
What to prepare
- An authorized original image.
- A short edit brief and a list of features that should stay stable.
Image record
Google Vertex AI · gemini-3.1-flash-image · · 1280 × 859 px
Original generation prompt
Use case: illustration-story. Make ONE original finished comic page in a clean three-column by two-row grid, exactly SIX equally sized panels, narrow even ivory gutters and complete outer margins. The provided image is the character identity and wardrobe reference, not a page to copy. Preserve its adult botanist with warm brown skin, short dark curly hair, teal jacket, coral scarf, sand trousers, brown boots, and round mint helper robot with two dark eyes and exactly one short antenna. Quiet elegant ink-and-watercolor comic style, teal/coral/sage/ochre palette. Tell a very simple readable rooftop greenhouse planting story: panel1 wide view botanist and robot arrive at the greenhouse carrying a small terra-cotta pot; panel2 medium view the botanist places the pot on the wooden bench; panel3 close view gloved adult hands holding a seed above the pot; panel4 medium view botanist and robot water the pot with a small cream watering can; panel5 close view botanist and robot look at the tiny seedling; panel6 wide view both stand by the finished row of three potted seedlings in afternoon light. All people are adults; each scene has one botanist and one robot unless only hands are framed. The pot remains terra-cotta, the greenhouse stays structurally coherent, identity and outfit remain recognizable. No writing, numbers, logos, speech balloons, captions, lettering, illegible marks, third-party characters, UI, photorealistic environment or claims of tested consistency. Landscape 3:2.
View full artworkLimit the first change
Change one attribute so you can see whether the request was followed.
Protect identity and geometry
Compare the face, hands, silhouette and object shapes with the original.
Review the whole image
A correct jacket color is not enough if the background or character has changed unintentionally.
A useful edit is judged by both the requested change and the important details it preserves.
A practical starting point
Start with an authorized source
Choose a clear image you have permission to edit. Keep the original available for comparison.
Describe one controlled change
Name the change and the details that should remain stable, such as a person’s identity, the object shape or the framing.
Compare inside and outside the edit
Check the requested area and surrounding details. Review lighting, edges, anatomy and unintended changes before keeping the result.
What to inspect
Image record
Google Vertex AI · gemini-3.1-flash-image · · 1280 × 859 px
Original generation prompt
Use case: photorealistic-natural / product-mockup. Create a natural window-lit editorial PHOTOGRAPH of exactly TWO anatomically normal adult hands holding one landscape matte comic proof sheet above a real warm oak desk. Image1 supplied is the original adult botanist and mint robot reference sheet; image2 is the COMPLETE original SIX-PANEL greenhouse comic. The held paper must have landscape 3:2 proportions so Image2 fits WITHOUT rearrangement. Put the exact complete Image2 sequence on the held sheet as THREE columns by TWO rows, with a clean ivory outer margin: upper row arrival, pot on bench, gloved seed close-up; lower row watering, seedling reaction, final greenhouse view. Preserve all six distinct scenes in that order; DO NOT repeat watering, remove seed, reformat into two columns, add text or combine panels. Keep the original adult's teal jacket, coral scarf, dark curly hair, sand trousers and original mint round robot. Hold the paper by its two lower blank outside corners only; all artwork corners remain visible and hands cover no panels. The source character-reference sheet from Image1 rests flat farther back on the desk, partly visible beside the held print. Only the printed artwork is ink/watercolor illustration; hands, table, thin paper grain and room are a convincing photograph. No visible faces, booklet, screen, logos, titles, numbers, writing, fake UI, metrics or product packaging. Natural skin/five fingers, coherent optical perspective, subtle paper flex, quiet soft daylight. The completed hand-held story takes 75 percent of frame. Fictional lifestyle presentation only, never a customer photograph or research result. Landscape 3:2.
View full artworkPersonal Agent · Creative Intelligence research
From Personal Agent memory to better creative results
A Personal AI Agent uses permissioned, evolving context to help someone complete work over time. For comic creators, the challenge is not only remembering a character, but retrieving the right story state, controlling the next image and verifying a creator-approved result. CyaniModel explores the visual-workflow layer of this larger architecture; DeepStory, Agent Loop and Feedback Learning are distinct proposed responsibilities, not claims of a fully deployed personal agent.
Research direction
Research behind the workflow
Measure unintended changes
We are exploring local correction methods with explicit review of the edited region and the scene outside it.
Check preservation as well as instruction following
An editor can follow one instruction while changing another important detail. Review should record both outcomes.
Research direction
Visual Review Cascade
Which checks can catch an identity or story error early, and which need a closer look?
How we would evaluate it: Track missed serious errors and incorrectly rejected panels alongside review cost and latency.
Research direction
Targeted Character Repair
Can one incorrect character detail be repaired while preserving the accepted scene?
How we would evaluate it: Inspect the edited area, the surrounding pixels and the story after each proposed repair.
Know the boundaries
- Exact preservation and region controls depend on the chosen editing tool.
- This planning case is not a measured before-and-after product test.
Primary research informs these questions; it does not establish results for CyaniModel.
Frequently asked questions
Can I edit with everyday language?
Describe the intended change clearly in a supported image-editing workflow. Check the available reference and editing options before starting.
How do I avoid changing the rest of the photo?
State the details to preserve, make a controlled first edit and compare the entire result. Text instructions alone do not guarantee exact pixel preservation.
Can I edit photographs of real people?
Use material you are authorized to edit and share. Review identity, context and the intended use carefully; the illustrative case here uses an original fictional character.
Continue your story
Use an image as the start of a new visual draft
Plan an image-to-image transformation with an authorized source. Decide which structure to preserve, what to change and how to review the resulting draft.
Generate an image with a clear creative brief
Plan an original image from a focused brief. Choose a subject, composition and visual direction, then inspect the result for its intended use.
Prepare a character for more than one image
Build a reference-led plan for recurring original characters. Record identity, outfit and useful views before reviewing new poses and scenes.
Make the intended change and the protected details explicit.
Continue in LlamaGen to check the available tools, model choices, account requirements and credits before generating.