How to generate images with Google Gemini: tools, limits, API access, and pricing explained.
- Graziano Stefanelli
- Sep 11
- 4 min read
You can create images directly inside Gemini chats, Workspace apps like Slides, or through the Gemini API.
In September 2025, Google Gemini offers one of the most accessible and integrated image generation pipelines in the consumer and enterprise AI space. All accounts—free or paid—can create images using Gemini’s underlying Imagen 4 model, internally codenamed Nano Banana, which is optimized for speed and stylistic diversity.
Users can describe a scene in plain English, such as “A neon koi pond with city reflections at night,” and receive AI-generated images in seconds. While the free plan includes limited daily access, paid plans offer significantly higher generation capacity and access to API controls for inpainting, editing, and resolution tuning.
Gemini’s image generation tool is available on web, mobile, and desktop for both free and Pro users.
The Gemini app interface on web, Android, and iOS includes a prominent “Image” button or camera icon. When a prompt is detected to contain visual intent (e.g. “Draw a castle at dusk in Van Gogh style”), Gemini responds with two images by default. Free users can do this approximately 10–30 times per day, depending on system load and availability.
Each image:
Is 1024×1024 pixels in PNG format
Includes SynthID invisible watermarking and a visible “Created with Gemini” label
Takes around 5 to 10 seconds to render
Is stored inline in the chat and can be downloaded individually
Once the daily generation quota is exceeded, the button grays out and a banner informs users they’ve reached the cap. The system resets each day at midnight UTC. On Pro and Ultra plans, the limit is instead managed via a monthly pool of AI Credits.
Inside Google Slides, Gemini can generate or expand custom images without leaving your presentation.
Gemini’s integration with Google Workspace Slides allows you to generate or expand images directly inside a presentation. Once enabled (for Gemini for Workspace or Labs users), the sidebar shows a prompt input:
“Generate a minimalist green background with floating leaves”
“Create an illustration of a team working remotely”
Gemini inserts the generated image into the slide canvas in real time. You can also select an existing image and choose “Expand background” to fill white space or adjust margins using AI-assisted outpainting. This tool is only active for paid Workspace tiers: Gemini Business, Enterprise, Education Premium, or Labs access.
Gemini’s API allows for image creation, editing, and style control using Imagen 4.
Developers and advanced users can create and edit images via the Gemini API, using the following endpoints:
Endpoint | Function | Required Input |
generateImage | Generates a new image from text | Prompt (≤800 chars), model name |
editImage | Edits an existing image based on mask and prompt | PNG + mask + prompt |
The most common model used for generation is gemini-2.5-flash-image-preview, optimized for quick returns and low latency.
Key features supported in the API:
Generate up to 4 images per request using n=4
Request stylistic hints (e.g. anime, watercolor, photorealistic)
Perform inpainting using masks to replace specific parts of an image
Expand image resolution to 2048×2048 or apply square-to-landscape variants (beta)
Control output format as base64 JSON or binary PNG
This API is widely used in enterprise dashboards, generative UX prototypes, and low-code platforms seeking image automation workflows.
Adobe Firefly and Express now route some image generations through Gemini Flash.
In August 2025, Adobe Firefly and Express began incorporating Gemini 2.5 Flash Image as one of the engines powering their “Quick Image Draft” and “Concept Preview” tools. This model provides:
Up to 20 free image generations per month for standard Adobe users
Unlimited access for Creative Cloud subscribers with Firefly integration
Output fully compatible with Express’s background removal, recoloring, and layout tools
While Adobe Firefly continues to use its own Firefly 3 model for high-fidelity photorealism, Gemini’s Flash variant is preferred for stylized drafts and instant content blocks.
Pricing for Gemini image generation depends on your plan or API usage.
Gemini pricing is now aligned around a token-based billing system rather than flat per-image pricing. Each 1024×1024 image consumes about 1,290 tokens, and the rate across models is calculated as follows:
Plan | Monthly AI Credits | Approx. image count (1024²) | Estimated cost per image |
Free | ~100–200 internal cap | ~10–30 images/day | Free (limited) |
Gemini AI Pro | 1,000 AI Credits | ~775 images | ≈ $0.04 each |
Gemini AI Ultra | 25,000 AI Credits | ~19,000 images | ≈ $0.04 each |
Workspace Enterprise | Contract-based | Unlimited or admin-set | Custom pricing |
Extra features like inpainting, resolution expansion, and sequential editing may consume more tokens per image. Enterprises with high output needs are encouraged to use the API, which offers batch generation, usage tracking, and quota enforcement.
Gemini's image editing tools allow complex refinement, background expansion, and multi-step design workflows.
In addition to static generation, Gemini supports interactive editing of uploaded or generated images. Users can:
Mask a portion of an image (e.g., sky, object, text box)
Instruct Gemini: “Replace background with city skyline at night” or “Add shadow around the character”
Chain edits within the same chat for progressive iterations
Edits maintain stylistic and lighting consistency. The SynthID watermark remains intact across modified versions, ensuring image provenance is preserved. This workflow is ideal for creating reusable content blocks, social media thumbnails, and marketing visuals.
Summary: Gemini’s image generation tools are accessible, safe, and deeply integrated into Google’s ecosystem.
Tool | Who can use it | Key strength | Limitations |
Gemini Chat (web/mobile) | Free & paid users | Instant generation, no setup | 10–30 images/day for free users |
Gemini in Slides | Workspace paid plans | Direct image flow into decks | Not available in free tier |
Gemini API | Developers, paid plans | 4 images at once, full control | Pay-per-token usage |
Firefly & Express (Adobe) | Adobe users | Gemini + Firefly hybrid pipeline | Gemini use limited to quick drafts |
By September 2025, Gemini has become one of the easiest ways to create high-quality images for free, with optional upgrades for businesses, creators, and developers who need scale. Whether generating slides, blog visuals, UI drafts, or experimental art, Gemini’s image tools now rival ChatGPT’s and Midjourney’s—but with native support inside Google’s productivity suite and a well-documented API.
____________
FOLLOW US FOR MORE.
DATA STUDIOS
