Prompt Library / 3D Avatar / Icon Prompt / for Kandinsky
This is the Kandinsky version of the 3d avatar / icon prompt prompt: a friendly 3D avatar or app icon with a clean, centred composition. It is formatted the way Kandinsky takes prompts best. Free — copy it, replace the bracketed parts, and paste it in.
Push copies this prompt with a reference line naming your image, and downloads the image file — attach it in the model to complete the pair. This page stages the pair for you; it cannot upload into the model itself (sites cannot reach into each other), so the final attach happens there. The image is downscaled to at most 1024px and stored only in this browser, paired with this prompt. If browser storage is full, it is kept for this visit only.
Kandinsky is Sber's open image/video family from Kandinsky Lab, published as open weights (the current generation is the Kandinsky 6 line) and also met through GigaChat. It takes natural-language prompts in Russian or English — write the scene as plain descriptive sentences: subject, environment, action, light. Being open weights, it runs locally with the usual diffusion controls (seed, steps, guidance) in your interface; the words here are the descriptive core.
Cost structure: open weights — the Kandinsky family runs on your own hardware, so the marginal cost per generation is zero; GigaChat is the hosted route. The same prompt is pre-configured in this library for every major model, so you can run it on whichever route fits your budget — comparing the structures before you generate is a free way to save money.
Replace the bracketed placeholders — [CHARACTER OR OBJECT], [COLOR SCHEME], [EXPRESSION OR ANGLE] — with your own details before pasting.
See all model versions of this prompt, or open the prompt gateway to build a custom prompt from your own one-line goal, formatted for Kandinsky or any other model.


Mpromp2 — the AI Prompt Generator. Every prompt in this library is free to use, including commercially.