Can I generate from text and edit with reference images?

Yes. The current Toono AI workspace supports text-to-image and reference-led image-to-image creation with Nano Banana 2. Use references when a person, product, animal, palette, material, or layout needs to shape the next result.

How many reference images can I add?

The current workspace accepts up to 14 reference images. That is a limit, not a target: a smaller set with one clear role per image is usually easier to direct and review.

Which resolutions, aspect ratios, and formats are available?

The current integration offers 1K, 2K, and 4K output, a range of square, portrait, landscape, and wide ratios, plus JPG, PNG, or WebP export. The active controls and credit estimate appear in the workspace before generation.

Are text and reference consistency guaranteed?

No. Nano Banana 2 can create convincing short text and reference-led edits, but spelling, identity, geometry, object count, and fine detail can still change. Review the full-resolution output and keep critical copy editable.

Fast image creation for everyday campaign rounds

Turn one idea into the next useful campaign image

Start from a prompt, add reference images when the subject or style needs an anchor, then create or revise the next asset without rebuilding the brief. Nano Banana 2 balances speed, visual quality, reference-led editing, short text, and everyday production flexibility in one workspace.

Create with Nano Banana 2

Open image generation with this model selected

See examples and review points

Reference editing
Short text
1K–4K

A botanical drink can, weekend market scene, terrier, color study, and poster direction form one campaign—not a gallery of unrelated samples.

Visual examples

Four production tasks to test with the same brief

The useful test is whether one subject survives a scene change, a targeted edit, a short-text layout, and a small campaign series. Each example below answers a different production question.

Google

Move the product into a believable market scene

Place the same slim can in a contemporary weekend-market setting, then inspect scale, condensation, contact, shadows, and whether the scene still reads clearly at feed size.

Google

Build a short-text poster direction

Generate a bold SATURDAY MARKET layout, then verify every letter, safe area, hierarchy, contrast, and relationship between the headline and the product.

Google

Change one campaign condition at a time

Use references to revise the table, color mood, dog position, or surrounding props while protecting the can silhouette, palette, camera angle, and core composition.

Google

Carry one subject across a compact image set

Compare the can and terrier across studio, market, picnic, and poster placements to find where shape, markings, material, or proportion begins to drift.

Examples show creative directions, not guaranteed replicas. Reference fidelity, spelling, object details, people, credit use, and generation time vary with the source material and selected settings. Review every output before publishing.

Practical workflow

Keep the brief small enough to revise

Nano Banana 2 is most useful when each round has one clear purpose. Lock the subject, define the placement, and change only the condition you are evaluating.

01
Write the non-negotiable details
Describe the subject, materials, palette, camera view, required short text, and anything that must not change. Add only the references that support those conditions.
02
Choose the real placement before rendering large
Set the aspect ratio for the feed, store, article, or banner, then prove the direction at a practical size before moving to a higher-resolution finish.
03
Review the round against the same checklist
Compare text, identity, product geometry, hands, small objects, shadows, crop, and export format. Revise one structural problem instead of rewriting the whole prompt.

Review checklist

A fast result still needs a production review

A polished thumbnail can hide a changed label, extra object, broken hand, or incorrect character. Check the constraints before deciding that a direction is finished.

Subject and reference fidelity
Compare silhouette, color, markings, face, clothing, product parts, materials, and repeated features with the sources at full size.
Text and information
Read every word, number, punctuation mark, and line break. Add critical prices, dates, claims, and legal copy later in a controlled design tool.
Composition and physical logic
Inspect scale, perspective, contact points, reflections, shadows, fingers, duplicate props, and whether the requested edit leaked into protected areas.
Delivery and permissions
Confirm dimensions, crop, file format, accessibility, reference-image rights, likeness permission, and the destination platform's current rules.

Nano Banana 2 workflow questions

Image inputs, output choices, and review points in the current Toono AI integration.

Compare more control or choose a guided image App

Try Nano Banana Pro for more complex reference assembly, compare GPT Image 2 for another image workflow, or open the AI Image Generator when you want to evaluate multiple models.