Nano Banana 2 creates new images from text and edits prepared visuals with up to fourteen reference images. Describe the subject, environment, materials and lettering for a new composition, or assign the layout, identity and material palette to their respective references before describing the change. Choose 1K, 2K or 4K and one of ten aspect ratios. When the brief needs external information, the task also offers optional Google Search assistance. Review the current resolution-based estimate and inspect the result against your source material and intended use.
Nano Banana 2 is developed by Google. Vidgo provides access to the model. Model names and trademarks belong to their respective owners.Choose Text to Image for a written scene and Image Editing for one to fourteen source images. Describe the sources, intended composition and details to preserve, then choose resolution and framing. Enable Google Search when external information is part of the creative brief and review the task estimate before generating.
Google's next-gen image model powered by Gemini 3.1 Flash with native 2K/4K resolution, chain-of-thought reasoning, precise multi-language text rendering, and up to 14 reference images.
Google's next-gen image model powered by Gemini 3.1 Flash with native 2K/4K resolution, chain-of-thought reasoning, precise multi-language text rendering, and up to 14 reference images.
The text task uses prompt with resolution, optional size and optional google_search. The editing task also accepts image_urls. One reference can define the exact room layout, while another supplies the wood, fabric, metal and wall finishes. Explain those roles and their relationship in the prompt, then identify the camera and proportions to preserve in the completed image.
| Model / task | Input and workflow |
|---|---|
| Nano Banana 2 Text-to-Image | Google's next-gen image model powered by Gemini 3.1 Flash with native 2K/4K resolution, chain-of-thought reasoning, precise multi-language text rendering, and up to 14 reference images. |
| Nano Banana 2 Image Edit | Google's next-gen image model powered by Gemini 3.1 Flash with native 2K/4K resolution, chain-of-thought reasoning, precise multi-language text rendering, and up to 14 reference images. |
Describe the composition and assign a clear purpose to each source. The examples and illustrations below cover a night landscape, a room layout, a material palette, their finished combination and the optional search-assisted workflow.
Specify the viewpoint, light, subject and environmental detail. The existing bay example asks for bioluminescent water around a wooden boat beneath a starry sky. Inspect the actual composition, reflections and light relationships when developing a landscape concept.
Provide the spatial reference and name the window, bench, bookcase, chair, table and rug positions to preserve. The reading-room source establishes the camera perspective and furniture proportions before the prompt adds material direction.
Use a palette image for finishes and explain which surface receives each material. The reading-room example assigns pale oak, limewash, rust-red boucle, bronze, green wool and cream linen to specific parts of the scene. Compare those roles with the generated surfaces.
Describe the shared camera, proportions, light and contact so the sources form one room. The existing Vidgo example combines the drawing and palette into a photographic reading space. Inspect furniture placement, material texture and daylight against both inputs.
Enable the optional Google Search setting and describe the information that should guide the image. Use the resulting visual together with the relevant source material when checking names, context and depicted details. The accompanying image is the existing task’s illustration of this workflow.
The first image comes from an existing public example of the same model and its original night-bay prompt. The second example is the Vidgo task’s reading-room demonstration: one reference supplies layout and the other supplies the material palette. Read the complete prompts and compare the room inputs with the displayed result to inspect furniture placement, camera perspective, finishes and daylight.
An aerial view of a bioluminescent bay at night, glowing blue-green plankton tracing the wake of a small wooden boat, stars reflected in the calm water, long exposure effect, the Milky Way arching overhead, National Geographic cover quality
Create a photorealistic finished reading room by combining the two references. Use Image 1 as the exact spatial and furniture layout: wide back window, low built-in bench beneath it, tall left bookcase, rounded lounge chair at right, circular side table, and centered rectangular rug. Use Image 2 as the exact materials palette: pale oak for the floor and bookcase, warm limewash walls, rust-red boucle on the chair, dark bronze on the side table, moss-green wool for the rug, and cream nubby linen for the bench cushions. Preserve Image 1's camera perspective and proportions. Add realistic daylight and material texture, but no extra furniture, no people, no text, and no advertising.
Use text for a new scene or prepare references for a room, subject, material palette or local edit. Assign the sources their roles, state the intended result and select resolution and framing. Inspect the completed image for a landscape concept, interior proposal, material study, character presentation or an integrated creative workflow.
Write the viewpoint, subject, light and environment, then choose resolution and framing. Inspect the generated spatial relationships and mood for a storyboard or visual proposal.
Provide the layout reference and identify the architecture and furniture positions to retain. Describe the finishes and light, then compare the image with the plan for an interior presentation.
Prepare a palette and a scene reference, assigning each material to its intended surface. Inspect texture, color and light in the result for a material study.
Supply clear references and describe the object, region or environment to change. State the identity, structure and camera to preserve, then review the edit for a character or scene concept.
Describe the external context relevant to the image and enable the supported Google Search option. Review the depicted details with the supporting material before selecting the result for a presentation.
Retain the selected task, prompt, image URLs, resolution, size and search setting. Submit the authenticated request and track its task ID to retrieve the completed image.
Nano Banana 2 uses the current per-generation price for the selected 1K, 2K or 4K resolution. The dynamic table shows the lowest current generation rate for each task. Open the text or editing task, configure the actual resolution and inputs and review the complete estimate before submitting. Use that amount and the planned number of tests to budget the project.
| Model | Provider | Workflow | Input material | Output specification | Starting rate |
|---|---|---|---|---|---|
| Nano Banana 2 Text-to-Image | Text to Image | Text prompt | 4K | $0.025 / generation | |
| Nano Banana 2 Image Edit | Image Editing | 1–14 reference images | 4K | $0.025 / generation |
Test your creative brief online, then bring the verified request into your application.
Enter the prompt and upload the editing references, using JPEG, PNG or WebP images within 10 MiB each. Select resolution and framing, set Google Search where appropriate and review the estimate. Generate the image and compare appearance, layout, materials and requested changes with the sources.
Choose a task, inspect its settings and cost, and follow the result from an online test to an authenticated API request.
Choose Text to Image for a written scene or Image Editing for one to fourteen references. Describe the sources, intended change and details to preserve. Choose 1K, 2K or 4K, a supported size ratio and the optional Google Search setting when the brief needs it.
Open the matching task, enter your brief and supply the required images. Review the generated result against subject appearance, composition and the requested changes before carrying the workflow into a project.
Select the actual resolution and inspect the task’s complete generation estimate before submission. Use that request cost and the planned number of tests when budgeting a composition or edit.
Read the actual prompt and inspect its source images and result together. Use those examples to understand how to assign input roles and describe the details to preserve or change in your own request.
Retain the task ID returned by the API to query status and retrieve completed file URLs. Use your online generation history to review earlier results and the creative direction you tested.
Use the documentation linked from each task for its exact model identifier, accepted inputs, authentication and response format. Integrate the same task you tested online so its media and settings match the request.
Nano Banana 2 is Google’s Gemini 3.1 Flash image model available on Vidgo through text generation and editing. It supports 1K, 2K and 4K, ten aspect ratios, editing with up to fourteen references and optional Google Search assistance. Use the prompt and source roles to define the composition and changes.
Explore model introductions, API guides and platform comparisons to understand input preparation, request integration and cost planning for your project.
