HOVIGuard

Image Generation

Create marketing images, mockups and concepts straight from the workspace -- without data leaving the EU.

Mobile workflow

01
Generate photo

A prompt produces a realistic photo.

02
Apply 'In suit' style

Pick a style preset — same face, new look.

03
Apply 'Comic' style

Same workflow, different style preset.

Three protection factors

EU models only

Image generation via EU-hosted providers -- no image uploads to US datacentres.

NSFW & IP filter

Before every generation a content and trademark check runs automatically.

Gallery & export

Generated images land in the workspace and can be exported or shared at any time -- 7-day default retention.

How image generation works

Five steps from prompt to finished visual -- transparent, EU-hosted, audit-proof.

  1. 1

    Write the prompt

    Describe the target image with style, setting, lighting mood. Concrete detail noticeably improves results.

  2. 2

    Pick model and resolution

    Gemini 3 Pro Image for detailed marketing visuals, Gemini Flash Image for quick iterations. Formats from square to 16:9 and 9:16.

  3. 3

    NSFW and copyright filters check

    Before generation the protection layers check the prompt: no NSFW content, no protected brands or persons. Violations are blocked.

  4. 4

    Generation in 5-30 seconds

    Depending on model and format generation takes 5 to 30 seconds. You see a live status and can derive a variant from any result.

  5. 5

    Gallery & export

    Images land in the workspace gallery. Default retention 7 days. Export as PNG, JPG or WEBP.

Templates

Four template categories — click instead of an empty prompt

Instead of prompting from scratch, pick a category and HOVIGuard supplies tested styles, resolutions and tonalities as a starting point.

Style

Photorealistic, watercolor, 3D render, comic, minimal line — one curated style per image with tested modifiers.

Business

Hero banners, portraits in CI colors, product mockups, pitch slides — ready templates for marketing & sales.

Entertainment

Birthday cards, social stories, creative avatars, event visuals — quick fun without long fiddling.

Background

Studio backdrops, branded textures, office ambience, abstract loops — instantly usable as slide or video background.

Prompt assistant

Guided questions — the most out of your idea

You bring the rough idea — the assistant asks for what's missing: framing, tonality, mood, style reference, target medium, resolution. Two sentences become a precise prompt that actually delivers.

Targeted follow-ups
3-5 questions, max. 30 seconds — ready to generate.
Better hit rate
Fewer re-generations, fewer tokens burned.
Combines with templates
Pick a template, the assistant fine-tunes the rest.
GDPR-aware
Trademark and person filters run before generation.

Who benefits from image generation?

Four roles for whom HOVIGuard replaces external image tools, stock photo subscriptions and design studio briefings.

Marketing & social media

Hero images for landing pages, social media posts, ad visuals -- no stock photo licences.

Sales & presentations

Slide visuals, concept mockups for pitches, individual sales materials -- in brand CI.

Product & design

Product visualisation, variant exploration, pre-concepts for renders -- before commissioning the design studio.

Education & training

Teaching materials, explanatory images, onboarding graphics -- straight in brand style.

Benefits

Your prompts do not train image models -- contractually guaranteed
Marketing visuals, mockups, concepts -- without external image tools
Images stay documented in the audit trail
Included in the Pro plan's fair-use budget -- model costs without markup

Frequently asked questions on image generation

Answers on models, resolutions, data protection, brand protection and commercial use.

Which AI models are available for image generation?+

Currently the Google Gemini image models (Nano Banana family): Gemini 2.5 Flash Image, Gemini 3.1 Flash Image, Gemini 3.1 Flash Lite Image and Gemini 3 Pro Image -- all via EU gateways on Vertex AI in EU regions, with data processing agreement (DPA).

Are my generated images used for model training?+

No. All providers have contractually committed not to use prompts or generated images for training. Agreements are documented in the DPA.

Which resolutions can I generate?+

All models deliver around one megapixel: 1024x1024 square, 1344x768 (16:9) for hero visuals, 768x1344 (9:16) for social media stories, plus 4:3 and 3:4.

How long are generated images stored?+

Default retention is 7 days. Deletion rules are transparently shown per image — you can export images any time before they expire.

How are brand, person and copyright protection implemented?+

Before every generation the MultiLayer Data Shield runs: NSFW filter, brand detection (logo, slogan), person-of-interest detection (politicians, celebrities). Violations are blocked with reason.

What does an image generation cost?+

Image generation runs on your plan's fair-use budget. Every generation is shown as a percentage of the budget -- model costs are passed through without markup.

Can I use generated images commercially?+

Yes, per the respective model licence. Since the brand and person filter blocks problematic content before generation, the produced images are legally clean to use.

How can I maintain the same style across multiple images?+

With a reference image (image-to-image), style templates and consistent prompt wording on a fixed model -- the typical branding setup for marketing teams.

Generate images

Start the demo and generate your first image right inside the workspace.

Start Free Trial