Krea v2 Medium
Krea v2 Medium is a creatively focused image model built from scratch for expressive aesthetics, advanced style transfer, and full creative control — with 1K resolution cap.
Overview
What is Krea v2 Medium
Krea v2 Medium is Krea AI's foundation text-to-image model, released in May 2026 and built from scratch to prioritize aesthetic diversity and creative control. It excels at expressive illustration, style transfer, and reference-guided generation, with support for tunable creativity and moodboards, priced at $0.04 per image on Venice.
Running it privately on Venice
On Venice, Krea v2 Medium runs with anonymized privacy — your prompts are not stored or profiled, ensuring private creative exploration. Use it for expressive, style-rich outputs without building a traceable history, and pay only per generation with no subscription lock-in.
Assessment
Strengths and limitations
- Exceptional aesthetic diversity: designed to avoid the 'AI look' and support artistic, expressive styles.
- Most advanced style transfer system available, allowing precise application of reference image styles.
- Tunable creativity parameter lets you control the level of aesthetic interpretation and visual richness.
- Supports moodboards and generative sliders for consistent, brand-aligned outputs.
- Optimized for illustration, anime, painting, and editorial-style visuals.
- Capped at 1K resolution: cannot generate 2K or 4K outputs, limiting use for high-fidelity deliverables.
- Not open weights, so self-hosting, fine-tuning, or LoRA training is not supported on this variant.
- Slower than turbo variants, with no real-time generation for rapid ideation.
Samples
Sample outputs
Generated on Venice with our standard prompt suite — the same prompts we run through every model of this type, so you can judge it like-for-like.
A retro travel poster with the bold headline "VENICE" in large condensed serif type, sunset color palette, clean layout
Photorealistic close-up portrait of a weathered fisherman at golden hour, 85mm lens, shallow depth of field, natural skin texture
A small red cube balanced on top of a large glossy blue sphere, with a green cone to the right, plain light-grey studio background
Cozy watercolor illustration of a hillside village in autumn, warm tones, soft paper texture
Capabilities
What it supports
- Text to image
- Image to image
Specifications
Datasheet
- Maker
- Krea AI
- Released
- May 12, 2026
- Modality
- Text-to-image, style transfer, moodboards, reference-guided generation
- Max resolution
- 1K (1024×1024)
- Open weights
- No — proprietary
- Aspect ratios
- 1:1, 3:2, 16:9, 2:3, 9:16
- Prompt limit
- 5,000 chars
- Privacy on Venice
- Anonymized — prompts not stored
- Available on Venice since
- May 2026
- License
- Proprietary
API
Call it from your code
Venice exposes this model through the REST API. Queue a generation with the model id.
curl https://api.venice.ai/api/v1/image/generate \
-H "Authorization: Bearer $VENICE_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "krea-v2-medium",
"prompt": "A serene mountain lake at dawn, photorealistic"
}' --output image.pngPricing
What it costs on Venice
Flat per-image pricing on Venice: $0.04 per generation.
New Venice accounts include a free daily allowance and 500 welcome credits — no credit card required.
Alternatives
How it compares
| Model | Max resolution | Strongest at | Open weights | Price (Venice) |
|---|---|---|---|---|
| Krea v2 Medium | 1K | Style & creative control | No | $0.04 / image |
| Grok Imagine High Quality (SOTA) | 2K | Photorealism | No | from $0.06 / image |
| Flux 2 Max | 4MP | Photoreal detail | No | $0.09 / image |
| Lustify v8 | — | Speed & NSFW | No | $0.01 / image |
Best for expressive illustration and precise style transfer with tunable creativity.
Use cases
What it is good for
- 01Editorial and expressive illustration where style and mood are central.
- 02Brand-aligned visual campaigns using moodboards and style references.
- 03Concept art and character design with consistent aesthetic direction.
- 04Creative exploration with controlled levels of abstraction and interpretation.
- 05Social media content requiring distinctive, non-generic AI aesthetics.
Prompting
Getting better results
Use detailed style references or moodboards to guide aesthetic direction.
Adjust the `creativity` parameter to balance fidelity and artistic interpretation.
Leverage aspect ratio support for platform-specific formats like 9:16 for mobile.
Keep prompts under 5,000 characters and include specific artistic references (e.g., 'in the style of ') for best results.
Version history
Current — built from scratch for creative control
FAQ
Frequently asked questions
Krea v2 Medium is Krea AI's foundation image model, released in May 2026 and built from scratch for creative control, expressive aesthetics, and advanced style transfer. It's optimized for illustration, anime, and artistic styles with tunable creativity and reference-guided generation.
On Venice, Krea v2 Medium costs $0.04 per image generated. Upscaling to 2× costs $0.02 and to 4× costs $0.08, though the base resolution is capped at 1K.
No, Krea v2 Medium is not free. It is available on Venice at a flat rate of $0.04 per image, with no free tier or credits included. Access requires payment per generation.
No. Krea v2 Medium is a proprietary model developed by Krea AI. While Krea 2 RAW is available under Apache 2.0 for research, the Medium variant used on Venice is closed and not open weights.
Krea v2 Medium supports up to 1K resolution (1024×1024) across aspect ratios like 1:1, 3:2, 16:9, 2:3, and 9:16. It does not support 2K or 4K outputs.
Yes, Krea v2 Medium features the most advanced style transfer system available, allowing you to extract and apply the visual style from reference images with high precision, making it ideal for brand-consistent or artistically coherent outputs.
Krea v2 Medium excels in creative control, style transfer, and expressive illustration, while Flux 2 Max leads in photorealistic detail and higher resolution (4MP). Choose Krea for artistic direction, Flux for realism and fidelity.
Run Krea v2 Medium privately
No prompt logging. No data used for training.