# Zeva

> Zeva is an AI image generation platform that unifies six leading models (Z-Image Turbo, GPT Image 2, Nano Banana 2, Seedream 4.5, Seedream 5.0 Pro, Grok Imagine 2.0) behind a single interface. Generate images from text prompts (text-to-image) or transform existing photos (image-to-image), with no design skills required.

Zeva is built for creators, designers, marketers, and indie makers who need fast, high-quality AI image generation without juggling multiple subscriptions or learning model-specific quirks. We handle the model selection complexity so you can focus on the result.

Zeva is an **independent product** — not affiliated with, endorsed by, or sponsored by OpenAI, Google, Alibaba, ByteDance, or xAI. All model names and trademarks belong to their respective owners. We access these models through the [fal.ai](https://fal.ai) platform.

## Products

- [Text to Image](https://z-image-ai.run/text-to-image): Generate images from a text description. Choose between 6 AI models with different speed/quality/cost trade-offs. 100 free credits on signup.
- [Image to Image](https://z-image-ai.run/image-to-image): Transform an existing photo using an AI model. Best for style transfer, identity-preserving edits, and refinement.
- [Pricing](https://z-image-ai.run/pricing): Credit-based pricing. One-time packs ($14.90 / $34.90 / $78.90) or recurring subscriptions (monthly $9.90+ / yearly $94.90+). All plans include text-to-image, image-to-image, no watermark, commercial use.
- [Prompt Examples](https://z-image-ai.run/prompts): Curated prompt library — 30+ ready-to-use prompts across photography, illustration, design, 3D render, and concept art.
- [How to Use](https://z-image-ai.run/how-to-use): Three-step workflow guide — describe or upload, choose your settings, generate and download.

## Image tools

Single-purpose tool pages built on the same image engine — pick a task, upload a photo, and get the result with no model choice and no prompt-writing skills (at most a one-line detail, like describing the new background). More tools are added regularly.

- [AI Image Tools (hub)](https://z-image-ai.run/image-tools): Directory of free AI photo tools — Ghibli & cartoon styling, action figures, Polaroid photos, background changing, old photo restoration, AI professional headshots, AI face rating & attractiveness test, looksmaxing glow-up, image-to-prompt extraction, and more added regularly.
- [Ghibli AI Generator](https://z-image-ai.run/image-tools/ghibli-generator): Turn any photo into Studio Ghibli–style art in seconds — soft hand-painted textures, warm pastel tones, faces stay recognizable. Free to start, no prompt needed, no watermark, download as JPEG/PNG. Not affiliated with Studio Ghibli.
- [AI Cartoon Generator](https://z-image-ai.run/image-tools/cartoon-generator): Turn any photo into clean, vibrant cartoon art in seconds — bold outlines, flat colors, faces stay recognizable. Free to start, no prompt needed, no watermark, download as JPEG/PNG.
- [AI Headshot Generator](https://z-image-ai.run/image-tools/ai-headshot): Turn a selfie into an AI professional headshot — studio lighting, business attire, your face kept recognizable. Optionally describe the style and background (e.g. corporate, navy suit, office blur). Free to start, no watermark, LinkedIn-ready JPEG/PNG.
- [Old Photo Restoration](https://z-image-ai.run/image-tools/old-photo-restoration): Restore old, damaged or faded photos — remove scratches, creases and spots, repair damaged areas, fix fading, and add natural color to black-and-white photos. Faithful and conservative, faces stay recognizable. Free to start, no watermark, download as JPEG/PNG/WebP.
- [AI Background Changer](https://z-image-ai.run/image-tools/background-changer): Replace the background of any photo with a scene you describe in one line — sunset beach, blurred office, plain white studio. The subject stays exactly as photographed, composited into the new scene with realistic lighting and clean edges. Free to start, no watermark, download as JPEG/PNG.
- [AI Face Rating](https://z-image-ai.run/image-tools/face-rating): AI face rating and attractiveness test in one — upload a selfie for a 1–10 rating with five sub-scores (symmetry, proportions, skin clarity, jawline, eye area), practical tips and a downloadable report card. Adults 18+ only; photos are analyzed in memory and never stored. Free to start with signup credits.
- [Looksmaxing AI](https://z-image-ai.run/image-tools/looksmaxing-ai): Upload a selfie and instantly see a realistic glow-up of your best self — clearer skin, well-groomed hair, a flattering outfit and better light — while keeping your face, hairstyle and glasses. One tap, no prompt. Doesn't rank you or use PSL/tier labels. For adults 18+ using their own photo. Free to start, no watermark.
- [AI Action Figure Generator](https://z-image-ai.run/image-tools/ai-action-figure): Turn a selfie into a boxed collectible action figure — blister-pack packaging, a name plate and themed accessories, with your face, hairstyle and outfit kept recognizable. Optionally name the figure and add accessories. Powered by Nano Banana 2. Free to start, no watermark, download as JPEG/PNG.
- [AI Polaroid Generator](https://z-image-ai.run/image-tools/ai-polaroid): Turn a selfie into an authentic vintage Polaroid — white instant-film frame, soft flash and faded retro tones, with your face and pose kept recognizable. Optionally add a scene or another person. Powered by Nano Banana 2. Free to start, no watermark, download as JPEG/PNG.
- [Image to Prompt Generator](https://z-image-ai.run/image-tools/image-to-prompt): Upload any image (or paste an image URL) and get a detailed, ready-to-use AI prompt — General, Midjourney, Flux or Stable Diffusion format (SD includes a negative prompt), in 16 output languages — with a six-field breakdown (subject, style, lighting, camera, palette, mood). Text in the image is quoted verbatim. Costs 60 credits per extraction (signup credits cover the first one; failures refunded); one click runs the prompt in the site's generator. Images are analyzed in memory and never stored.

## Models

- [Z-Image Turbo](https://z-image-ai.run/models/z-image-turbo): By Alibaba. Fastest model — generates in 2-5 seconds at $0.005 per image (5 credits). Up to 1280px output. Best for rapid ideation and full-scene rewrites.
- [GPT Image 2](https://z-image-ai.run/models/gpt-image-2): By OpenAI. Highest quality, three quality tiers — Low ($0.012, ~15s), Medium ($0.05, ~60s), High ($0.22, ~2min). Best for professional output, complex compositions, and accurate text rendering in images (signs, posters, labels).
- [Nano Banana 2](https://z-image-ai.run/models/nano-banana-2): By Google. Best for instruction-based edits ('change X to Y') and identity preservation. Resolutions from 1K to 4K. Pricing: $0.08 (1K) / $0.12 (2K) / $0.16 (4K).
- [Seedream 4.5](https://z-image-ai.run/models/seedream-4-5): By ByteDance. Photorealistic results with strong text rendering and subject consistency. $0.04 flat per image. Supports up to 10 reference images.
- [Seedream 5.0 Pro](https://z-image-ai.run/models/seedream-5-0-pro): By ByteDance. Flagship model (July 2026) — reasons before drawing. Best for dense infographics, structured layouts, and native text in 14 languages (English, Spanish, Arabic, Japanese, Korean and more). $0.0675 flat per image (70 credits), output up to 1920px. Supports up to 10 reference images. No seed parameter.
- [Grok Imagine 2.0](https://z-image-ai.run/models/grok-imagine-2-0): By xAI. Released August 2026; ranked #2 on LMArena for both text-to-image and image editing (behind GPT Image 2). Best for designer-grade typography, posters, precise instruction edits, and multi-format creative — 9 aspect ratios including ultra-wide 2:1 and tall 1:2, at 1K or 2K. 60 credits (1K) / 80 credits (2K); image-to-image (one reference image) costs 70 credits at 1K / 90 credits at 2K. No seed parameter; safe-for-work API (no NSFW mode).

## Blog

AI image generation guides built on real generations — we run the prompts, compare models with identical inputs, and publish the unedited results.

- [Blog](https://z-image-ai.run/blog): All guides, prompt techniques, and model comparisons.
- [Seedream 5.0 Pro Review: 5 Real Tests of ByteDance's New Model](https://z-image-ai.run/blog/seedream-5-0-pro-review): Hands-on review — dense infographics, bilingual poster typography, photoreal portraits, plus a same-prompt shootout against Nano Banana 2 and GPT Image 2 (all outputs generated on Zeva, one take per model).
- [Nano Banana Pro vs Nano Banana 2: Which Should You Use?](https://z-image-ai.run/blog/nano-banana-pro-vs-nano-banana-2): Comparison of Google's two Nano Banana tiers on 4K, speed, text, and price — Pro specs drawn from Google and independent tests, Nano Banana 2 shown with our own edits, plus how to run Nano Banana 2 free and watermark-free with no API key.

## Key features

- 6 AI models unified in one workspace — switch models without losing your prompt
- Text-to-image AND image-to-image in the same interface
- Up to 9 aspect ratios per model (1:1, 3:2, 2:3, 4:3, 3:4, 16:9, 9:16; Grok Imagine 2.0 adds 2:1 and 1:2)
- Output formats: JPEG, PNG, WebP
- Credit-based pricing — pay only for what you use
- 100 free credits on signup, no credit card required
- No watermark, full commercial use rights on all paid plans
- Server-side content moderation (Creem Moderation API + per-model safety filters)
- Failed generations automatically refund credits

## Brand & company

- [About Zeva](https://z-image-ai.run/about): The team behind Zeva and why we built a unified AI image generator. Independent product, small team, not VC-backed.
- Contact: support@z-image-ai.run (typically responds within 1 business day)

## Legal & policies

- [Privacy Policy](https://z-image-ai.run/privacy-policy): How Zeva collects, uses, and protects user data. GDPR / UK GDPR / CCPA compliant. We do NOT train AI models on user data.
- [Terms of Service](https://z-image-ai.run/terms-of-service): User agreement, content ownership, billing terms, refund policy.
- Acceptable Use Policy: Covered in Terms of Service §3. No CSAM, no non-consensual sexual imagery, no deepfakes of real people, no hate speech, no illegal content. Violations are blocked pre-generation by Creem Moderation API and not eligible for refund.