Introducing Venice Image Engine V2

Venice provides private, uncensored generative AI. Today, we're introducing the Venice Image Engine V2, our most significant upgrade to image generation, combining exceptional quality with complete creative freedom.

Key Components of Venice Image Engine V2

Venice Image Engine V2 consists of two major components:

1. Venice SD35 (Default model): Custom-configured Stable Diffusion 3.5 engine powered by a Comfy UI workflow backend.

2. Upscale & Enhance (Pro Only): Completely new upscaling architecture with proprietary enhancement that creatively in-fills new pixels at the higher resolution.

You can try the new Venice SD35 model here for free.

1. Venice SD35: Custom-Configured Stable Diffusion 3.5

Venice implemented substantial improvements to the base Stable Diffusion 3.5 architecture:

• Natural language understanding: Venice SD35 processes conversational language more effectively, so you can describe images as you would to a human artist instead of using awkward keyword lists.

• Custom image generation pipeline: a specialized image generation Comfy UI workflow on the backend that delivers superior results through optimized processing techniques that standard implementations don't provide.

All these modifications mean Venice SD35 outperforms standard implementations in detail, creativity, and prompt adherence while maintaining your privacy.

The improvements in Venice Image Engine V2 are immediately apparent when comparing outputs:

• Superior photorealism: Venice SD35 produces images with notably more realistic textures, lighting, and details than previous models.

• Better prompt understanding: The new model follows complex instructions and natural language prompts more accurately.

• Reduced artifacts: Common issues like distorted faces, and unnatural textures are dramatically reduced.

• Enhanced artistic range: From hyperrealistic photography to stylized illustrations, Venice SD35 delivers exceptional results across diverse artistic styles.

Let’s compare the new default image model Venice SD35 to Fluently, our previous default image model.

You can try the new Venice SD35 model here for free.

Fluently (old default)	Venice SD35 (new default)

2. Upscale & Enhance (Pro Only)

The new upscaling system represents a significant leap beyond traditional pixel-by-pixel methods:

• Upscale: Doubles or quadruples your image resolution while preserving the original style and details.

• Better retention of fine details: Unlike traditional upscalers that simply duplicate pixels, our upscaler maintains crisp edges and textures.

• Fast rendering: Get high-resolution results in ~30 seconds.

Beyond simple upscaling, the new Enhance feature can transform your images:

• Detail amplification: Creatively adds realistic details that weren't visible in the original image.

• Latent space processing: Instead of just manipulating pixels, our enhancer works in the AI's conceptual space to generate authentic new details.

• Controllable: Adjust how much creative liberty the AI takes when enhancing your image, from subtle refinements to dramatic reimaginings.

How to Use Upscale and Enhance Features for Better AI Images

For the best possible results:

1. Use Upscale for accurate enlargements: Use the Upscale feature when you want to maintain the exact style and composition of your image while increasing resolution.

2. Use Enhancer for creative enhancements: Use the Enhance feature when you want to add details and refinements beyond what's in the original image. Adjust the creativity setting based on how much creative liberty you want the AI to take. For consistent results we suggest loading in your original prompt in the prompt field.

3. Combine for professional outputs: Combine both approaches by first generating with Venice SD35, then upscaling, and finally applying a subtle enhancement for the highest quality results.

Let’s compare zoomed-in screenshots for the following image.

Click right to scroll through to the Upscaled and Enhanced images:

Venice SD35

How to Get the Best Results with Venice Image Engine V2:

To get the most out of Venice Image Engine V2, consider these prompting strategies:

Be Detailed and Specific

Unlike previous models where adding too many details could hurt results, Venice SD35 thrives on detailed descriptions. Don't hold back, include specifications for:

• Subject details (appearance, pose, expression)

• Setting elements (location, time of day, atmosphere)

• Lighting characteristics (soft, dramatic, natural)

• Style references (photorealistic, fantasy art, oil painting)

Write in Natural Language

This is a big upgrade. Venice SD35 better understands natural, conversational language better than keyword lists. Instead of:

woman, red hair, forest, sunset, fantasy, detailed

Try something more natural:

A contemplative woman with flowing red hair walking through an ancient forest at sunset, fantasy style with intricate details and atmospheric lighting

Here’s a prompt template:

[Art Style/Photography Type], [Detailed Description of the Main Subject]. [Specific Features and Textures]. [Lighting and Environment]. [Additional Elements and Mood]. [Overall Style and Atmosphere]

Prompt examples

Hyper-detailed, 4K macrophotograph of a cat's eye. A mesmerizing close-up of its iris. The image captures every intricate texture and color within the iris: a stunning mix of golden amber, deep green, and flecks of copper radiating from the vertical, slit-shaped pupil. Fine details like the subtle ridges, reflections of light, and micro-textures on the eye’s surface are sharply visible. The glossy surface reflects a soft window light or a natural outdoor scene, adding depth and realism. Around the eye, ultra-fine fur in shades of grey and black fades softly into blur, emphasizing the hypnotic beauty and intensity of the feline gaze. The overall mood is captivating and intense.

Fantasy-style pencil sketch of a beautiful bumblebee resting on an otherworldly purple flower. The bee is illustrated with fine, detailed linework — its furry body slightly exaggerated in texture, with elegantly veined wings that shimmer faintly with magical symbols etched into them. The flower is large and ornate, its petals curling like tendrils, with a glowing core and faint mystical runes along the stem. Ethereal wisps and tiny spark-like particles float around the scene, adding a sense of enchantment. The background includes faint outlines of fantastical foliage and faint traces of faerie lights, drawn in soft, swirling strokes. The overall style is whimsical, detailed, and dreamlike — as if plucked from the pages of a magical creature’s sketchbook.

Hyperrealistic cyborg, with intricate circuitry and weathered metallic skin, is captured in a close-up portrait, revealing the fine lines and minute details of its advanced synthetic features, bathed in the cold, sterile light of a futuristic laboratory. The cyborg's optic sensors glow with a faint, eerie luminescence, hinting at the complex emotions and data processing occurring beneath its mechanical exterior. The image is rendered in a meticulous, almost photographic style, emphasizing the stark contrast between the organic and artificial elements of the cyborg's design. The overall mood is eerie and futuristic, evoking a sense of advanced technology and mystery.

Enhance Dos & Don’ts

For Enhancing, follow these tips for additional prompting

• DO use style prompts: Anime or Photography or Sketch to change the overall look of your image. With high hallucination this works well.

• DO use material prompts: Metallic, Stone Surface, Wood Surface to change the texture of something in your image

• DON’T use subject prompts: adding additional subjects (animal, body parts, people etc) in your prompt will lead to unexpected results

Best practice for all around enhancement is using the original prompt that generated your image.

Enjoy Stunning Creative Freedom with Venice Image Engine V2

Venice Image Engine V2 represents our continued commitment to providing powerful, unrestricted AI capabilities while maintaining complete privacy. Your prompts and images remain private, stored only in your local browser—never on our servers.

You can try the new Venice SD35 model here for free.