---
source: https://autoadify.com/ai-models/krea-2-medium
title: "Krea 2 Medium for social media"
---

# Krea 2 Medium for social media

> Krea's balanced model, known for stable, consistent results and tagged for both photoreal and illustration. Useful when a series of posts needs to look like it belongs together.

Krea 2 Medium is one of 80+ AI models in Autoadify, the AI agent for social media. No Krea account or API key is needed: every model is in one subscription, chosen per generation — by asking the AI agent, in a social media AI workflow step, or in AI Studio.

- **Lab:** Krea
- **What it is:** Krea's balanced model — stable, consistent generations
- **Best for:** Photoreal, Illustration
- **Available on:** Pro plan and above
- **Cost band:** Mid, relative to its own type

## What to make with Krea 2 Medium

- **Product and lifestyle shots** (Instagram, Facebook): Images a customer judges a product by, for feed posts and ads.
- **Scenes for a launch** (Instagram, Facebook, LinkedIn): The product in a setting: a kitchen, a beach, a desk, without a photo shoot.
- **Explainer graphics** (LinkedIn, Instagram): Concepts and processes, where a photo would be the wrong answer.

**Pick it when:** You want a consistent look across a set of images.

**Try Krea 2 Medium Turbo instead** for faster, cheaper iteration.

## Sizes it will be posted at

The size each platform recommends for a feed image, so you know what to ask for before you generate.

| Platform | Spec | Note |
| --- | --- | --- |
| Instagram | 1080 × 1080 | Square Post: 1:1 — universally safe, no cropping risk |
| Facebook | 1200 × 630 | Feed Image Post: 1.91:1 — standard post, renders well on all devices |
| LinkedIn | 1200 × 627 | Post Image: 1.91:1 — standard feed post image |
| X (Twitter) | 1600 × 900 | Tweet Image: 16:9 — displays inline in feed, safe area centered |
| YouTube | 1280 × 720 | End Screen: 16:9 — added to final 5-20 seconds of video |

## Sample prompts

- **Instagram feed / ad** — prompt: "Editorial campaign photograph of a contemporary performance sneaker with a sculptural aerodynamic silhouette, suspended slightly above a brutalist concrete ledge with a soft contact shadow beneath it. Environment: underground urban architecture in raw brutalist concrete, geometric shadows, faint atmospheric haze. Dramatic late-afternoon sunlight cuts diagonally through the space. Low-angle hero composition on a 28mm lens, the product occupying roughly the central 45% of the frame, dynamic perspective, crisp material detail — layered technical mesh, matte synthetic panels, translucent sole elements, subtle reflective trim — a controlled sense of motion with no motion blur. Sophisticated monochrome architecture with restrained warm accents, cinematic contrast, subtle film grain. Generous empty space above the product for campaign typography to be set later. Unbranded: no logo, no monogram, no maker's mark anywhere on the product. No text, no watermark."

  Output: https://d2ngr02qv3hi28.cloudfront.net/preset-samples/v2/campaign-hero-shot-c73dde2f9e.jpg
- **Instagram feed post** — prompt: "Professional architectural photograph of a contemporary living room with floor-to-ceiling windows onto a lush green garden. Curved plaster walls and warm natural oak flooring, minimalist but warm, soft indirect lighting, natural stone and timber details. In frame: a sculptural cream sofa, a travertine coffee table, textured linen curtains, handmade ceramics, one large abstract artwork and a few plants — every object intentional and realistically scaled. Late-afternoon sunlight enters from the left, casting long soft shadows across the floor. Warm neutral cream, sand, natural wood and muted stone palette. 24mm wide-angle lens, vertical lines kept perfectly straight, true proportions, realistic material texture, subtle depth, high-end interior-design editorial, natural photographic lighting, photorealistic. No people, no text, no logos, no watermark."

  Output: https://d2ngr02qv3hi28.cloudfront.net/preset-samples/v2/interior-space-ed216a318b.jpg
- **Facebook / LinkedIn post** — prompt: "Premium lifestyle advertisement, photorealistic editorial photography, 16:9. A young woman in an oversized cream knit sweater and relaxed beige trousers, relaxed and genuinely at ease on a neatly made bed in a minimal beige bedroom, plants and a round mirror behind her, holding a plain white ceramic mug in both hands, looking slightly off-camera with a natural unposed expression. Warm morning light, soft natural shadows, shallow depth of field, realistic skin and fabric texture, warm neutral colour grading. The subject sits in the LEFT half of the frame; the right half stays calm and uncluttered so type can sit on it. Set across that right half, in large clean white serif type: Small / Steps / Every Day — one line per slash, the slashes themselves not drawn. Beneath it, smaller white sans-serif type reading exactly: Be proud of how far you've come. A single hand-drawn white underline swoops beneath the last line, and three small hand-drawn white hearts sit in the upper-right corner. Spell every character exactly as given. No logo, no watermark, no other text anywhere in the frame."

  Output: https://d2ngr02qv3hi28.cloudfront.net/preset-samples/v2/lifestyle-affirmation-a9d05b8f3f.jpg
- **LinkedIn / Facebook post** — prompt: "Photorealistic editorial still life, premium commercial photography. A wooden desk by a large window, a dark ceramic mug of coffee, a small stack of notebooks and a pen, green plants and open countryside beyond the glass, lit by soft natural morning light, background softly blurred. A matte-black card with slightly rounded corners and a soft shadow stands upright on the right of the frame, angled as a real object in the scene. Set on the card, in elegant highly legible type, exactly this and nothing more: "Discipline today leads to freedom tomorrow.

Keep showing up." — line breaks kept as written, large decorative quotation marks opening and closing it. The type is white, with the words "showing up" picked out in warm gold. Small minimalist type on the mug reads: FOCUS / PLAN / EXECUTE, stacked on three lines. Generous negative space, uncluttered composition, realistic wood, ceramic and paper texture, calm aspirational mood. Spell every character exactly as given and add no other text, no logo, no watermark."

  Output: https://d2ngr02qv3hi28.cloudfront.net/preset-samples/v2/quote-still-life-fdf86594fb.jpg
- **LinkedIn / X post** — prompt: "Premium educational infographic, 16:9, split composition on a deep navy and black with electric blue and subtle purple accents background. LEFT HALF, typography only: a small wide-tracked eyebrow line reading exactly PROMPT ENGINEERING, a thin rule beneath it, then a heavy three-line headline in bold condensed sans-serif reading exactly: TIPS TO GET / BETTER AI / RESPONSES — one line per slash, the slashes themselves not drawn, the middle line in the accent colour and the rest white. Below it, four evenly spaced rows, each a small rounded-square line icon in the accent colour beside a short label, reading exactly, top to bottom: Be Clear & Specific, Provide Context, Use Examples, Iterate & Refine. RIGHT HALF, photographic: a young professional in a dark hoodie at a laptop in a dark modern workspace, lit by the screen, thoughtful and mid-task. Around them, subtle futuristic interface graphics — a glowing neural-network constellation, faint data visualisations, thin connecting lines — and two floating rounded interface cards, the upper one headed INPUT PROMPT over the line "Explain the concept of quantum computing in simple terms.", the lower one headed AI RESPONSE over the line "Quantum computing uses qubits that can exist in multiple states at once...". Cinematic lighting, sharp detail, generous breathing room, premium SaaS marketing design. Spell every quoted string exactly as written. The interface is generic and invented — no real company's logo, product mark or name anywhere in the frame, no watermark."

  Output: https://d2ngr02qv3hi28.cloudfront.net/preset-samples/v2/tips-infographic-d7bd3014e4.jpg

## Alternatives

- [Krea 2 Medium Turbo](https://autoadify.com/ai-models/krea-2-medium-turbo): Distilled Krea 2 Medium — fastest iteration
- [Nano Banana 2](https://autoadify.com/ai-models/nano-banana-2): Google Gemini 3.1 Flash image
- [Nano Banana Pro](https://autoadify.com/ai-models/nano-banana-pro): Google DeepMind — 2K sharp imagery
- [GPT Image 2](https://autoadify.com/ai-models/gpt-image-2): OpenAI GPT Image 2 — text to image

---

_Source: https://autoadify.com/ai-models/krea-2-medium_