What is Image Generation?
AI image generation is the process of creating new images with machine learning models that have learned visual patterns from large datasets. Given inputs like a text prompt or a reference image, models synthesize pixels to produce photorealistic or stylized results. You can control outputs with settings such as style keywords, aspect ratio, number of steps, guidance strength, seeds, and by using techniques like image-to-image, inpainting (editing parts), and outpainting (extending beyond the original).
Typical uses include concept art, marketing visuals, product mockups, storyboards, and photo edits like background removal or upscaling. Results can contain artifacts and reflect dataset biases, so review outputs and respect copyright and usage policies. On Hatz AI, you can generate or upload images and iteratively edit them within Secure Chat, with the chat ref recent context to streamline revisions.
How Can I use Image Generation?
In Secure Chat, you can generate images from text prompts or upload existing images to edit—without leaving the conversation. The chat retains context and can reference up to the 3 most previous images, letting you iterate with follow-ups to refine style, colors, composition, or aspect ratio; request variations; and apply edits like upscaling, background removal, or object additions. Each version stays linked in the thread, making it easy to review, compare, download, and reuse outputs as you continue generating and editing in one place.
How do I know what model to pick for Image Generation?
The AI Model Selector will show Image Generation on models that have this capability
Where can I generate images in Hatz?
Currently Image Generation is supported in Secure Chat.
Which model is the best for Image Generation?
Best depends on the task, style, and quality bar. Use the AI Model Selector to choose a model with the Image Generation capability. For API requests, you can use auto-routing or pin an image-capable model when you need more predictable behavior.
How many generated images can a chat reference?
A chat can reference up to 3 previously generated images.
How can I save an image?
Generated images include controls to open, download, and expand the image preview. You can also use your browser's normal image save options after opening the image.
Google Chrome (Windows/Mac)
Save image: Right‑click the image > “Save image as…”, choose a folder, Save.
Get full resolution: Right‑click > “Open image in new tab,” then right‑click > “Save image as…”.
Copy image URL: Right‑click > “Copy image address,” or open in a new tab and copy the URL from the address bar.
Mozilla Firefox (Windows/Mac)
Save image: Right‑click > “Save Image As…”.
Get full resolution: Right‑click > “Open Image in New Tab” (or “View Image”), then save.
Copy image URL: Right‑click > “Copy Image Link” (older versions: “Copy Image Location”), or copy from the address bar after opening in a new tab.
Microsoft Edge (Windows/Mac)
Save image: Right‑click > “Save picture as…”.
Get full resolution: Right‑click > “Open image in new tab,” then save from the new tab.
Copy image URL: Right‑click > “Copy image link/address,” or copy from the address bar after opening in a new tab.
Safari (macOS)
Save image: Control‑click or two‑finger click > “Save Image to Downloads” or “Save Image As…”.
Get full resolution: Control‑click > “Open Image in New Tab,” then use File > Save As, or Control‑click > “Download Image”.
Copy image URL: Control‑click > “Copy Image Address.” If you don’t see it, open the image in a new tab and copy the URL from the address bar.
Will Generating Images use all of my credits?
In practice, Image Generation is treated similarly to uploading a file, so you can experiment with prompts and make iterative tweaks without seeing a significant impact on your credit balance.
Actual usage can vary based on your choices (for example, resolution, number of variations, or other advanced options). For the most accurate and up-to-date details on how credits are calculated, please see our Credit Usage FAQ
Can Hatz see the Images I Generate?
No. The images you generate are encrypted.
Can I generate an image via the API?
Yes. Use the Chat Completions API with model: "auto" for image-generation prompts, or pin an image-capable model such as gemini-3.1-flash-image-preview or gemini-3-pro-image-preview when available to the user. Generated image output is returned as an image link when Hatz can process the provider response. See the Chat Completions API docs.