inference · 15.2s

Example inputhover

seed: 2397720375
width: 1024
height: 768
prompt: "photo of old village, evening, clouds"
strength: 0.8
scheduler: "DPMSolverMultistep"
guidance_scale: 7
negative_prompt: "(deformed iris, deformed pupils, semi-realistic, cgi, 3d, render, sketch, cartoon, drawing, anime), text, cropped, out of frame, worst quality, low quality, jpeg artifacts, ugly, duplicate, morbid, mutilated, extra fingers, mutated hands, poorly drawn hands, poorly drawn face, mutation, deformed, blurry, dehydrated, bad anatomy, bad proportions, extra limbs, cloned face, disfigured, gross proportions, malformed limbs, missing arms, missing legs, extra arms, extra legs, fused fingers, too many fingers, long neck"
use_karras_sigmas: true
num_inference_steps: 30

Realistic Vision

Array·realistic-vision·by Stability

Realistic Vision generates lifelike images, ideal for creative and professional projects.

Try it now →

API reference

Runtime (p50): 10s
Estimated price: $0.00108 / sec

Call the API

prediction.sh

curl -X POST \
  -H "X-API-Key: $EACHLABS_API_KEY" \
  -H "Content-Type: application/json" \
  --data '{
    "model": "realistic-vision-v6-0-b1",
    "version": "0.0.1",
    "input": {
        "seed": 2397720375,
        "width": 1024,
        "height": 768,
        "prompt": "photo of old village, evening, clouds",
        "strength": 0.8,
        "scheduler": "DPMSolverMultistep",
        "guidance_scale": 7,
        "negative_prompt": "(deformed iris, deformed pupils, semi-realistic, cgi, 3d, render, sketch, cartoon, drawing, anime), text, cropped, out of frame, worst quality, low quality, jpeg artifacts, ugly, duplicate, morbid, mutilated, extra fingers, mutated hands, poorly drawn hands, poorly drawn face, mutation, deformed, blurry, dehydrated, bad anatomy, bad proportions, extra limbs, cloned face, disfigured, gross proportions, malformed limbs, missing arms, missing legs, extra arms, extra legs, fused fingers, too many fingers, long neck",
        "use_karras_sigmas": true,
        "num_inference_steps": 30
    },
    "webhook_url": ""
}' \
  https://api.eachlabs.ai/v1/prediction/

Documentation8 sections

Overview
realistic-vision-v6.0-b1 — Text-to-Image AI Model

Developed by Stability as part of the realistic-vision family, realistic-vision-v6.0-b1 is a cutting-edge text-to-image AI model that generates photorealistic images from text prompts, solving the challenge of creating lifelike visuals without expensive photography or rendering setups. This Stability text-to-image powerhouse excels in producing highly detailed, human-like portraits and scenes with natural lighting, skin textures, and proportions that rival professional photography. Ideal for creators seeking a text-to-image AI model for realistic outputs, realistic-vision-v6.0-b1 supports high-resolution generations up to 1024x1024 pixels, making it perfect for AI image generator for realistic photos workflows.
Capabilities
Photorealistic Image Generation:
Generates highly realistic images, suitable for professional and creative purposes.
Wide Range of Subjects:
Capable of producing images of human faces, full-body compositions, objects, and complex scenes.
High Customizability:
Allows users to fine-tune prompts, resolutions, sampling methods, and negative prompts for highly specific outputs.
High-Resolution Outputs:
Supports resolutions up to 1152x640 for full-body compositions and 896x896 for portraits, with options for high-resolution fixes.
Artistic Flexibility:
Compatible with artistic and professional projects, enabling users to create customized artwork or concept visuals.
Hyper-Realistic Visuals: Generates images with exceptional detail and realism.
Versatile Applications: Ideal for artistic, professional, and research purposes.
Scalable Performance: Efficiently handles diverse text-to-image tasks.
Use cases
Use Cases for realistic-vision-v6.0-b1

For content creators, realistic-vision-v6.0-b1 transforms simple prompts into stunning portraits; a photographer can input "middle-aged woman with freckles smiling in a sunlit park, ultra-realistic, 8k" to generate stock-ready headshots for portfolios, bypassing costly shoots. Marketers leverage its photorealism for e-commerce, feeding product descriptions like "sleek black smartphone on wooden table with steam from coffee mug nearby, natural indoor light" to produce compelling visuals that boost engagement without studios.

Developers integrating the realistic-vision-v6.0-b1 API build dynamic AI image generator for realistic photos apps, such as virtual try-on tools where users describe outfits on diverse body types for fashion retail. Designers in advertising use it for scene composition, creating "corporate executive in modern office shaking hands, motivational poster style with city skyline view" to prototype campaign assets quickly and realistically.
Tips & tricks
How to Use realistic-vision-v6.0-b1 on Eachlabs

Access realistic-vision-v6.0-b1 seamlessly through Eachlabs' Playground for instant text-to-image testing, API for scalable realistic-vision-v6.0-b1 API deployments, or SDK for custom apps. Input a detailed prompt, select resolution up to 1024x1024 and aspect ratio, then generate high-quality PNG/JPEG outputs with photorealistic fidelity in seconds—no complex setup required.
---
Technical spec
What Sets realistic-vision-v6.0-b1 Apart

realistic-vision-v6.0-b1 stands out in the competitive text-to-image landscape by prioritizing hyper-realistic human anatomy and environmental details, outperforming many models in rendering subtle facial expressions and fabric textures without artifacts. This enables users to generate professional-grade product mockups or character designs that pass as real photos, reducing post-processing needs. Unlike generic generators, it handles complex prompts with multiple subjects and lighting conditions seamlessly, supporting aspect ratios from 1:1 to 16:9 and PNG/JPEG outputs with average processing times under 30 seconds on standard hardware.
- Superior photorealism in portraits: Excels at lifelike skin tones, hair strands, and eye reflections, allowing realistic-vision-v6.0-b1 API integrations for e-commerce AI photo generator apps where authenticity drives conversions.
- Enhanced detail retention: Maintains intricate elements like jewelry or distant backgrounds in high-res outputs, empowering designers to create realistic AI images for advertising without manual refinements.
- Flexible prompt adherence: Interprets nuanced descriptors like "golden hour lighting on a rainy street" with precise realism, ideal for developers building Stability text-to-image tools for storytelling visuals.
Things to be aware of
Experiment with Different Styles:
Use prompts describing specific aesthetics like "cinematic lighting" or "modern minimalism."
Create a Unique Portrait:
Generate a detailed face portrait at 896x896 resolution with precise features and expressions.
Design a Scene:
Try a full-body or half-body composition set in a specific environment, such as "a bustling marketplace at sunset."
Use Negative Prompts:
Include terms like “cropped, low-quality, extra limbs” in the negative prompt field to remove artifacts and refine output.
Explore High-Resolution Fix:
Use the Hires.Fix option to upscale your favorite images and improve sharpness and detail.
Try Abstract Concepts:
Input a creative prompt such as "a surreal landscape with floating islands and glowing plants."
Adjust Sampling Parameters:
Experiment with different samplers like DPM++ SDE Karras and increase sampling steps for more refined images.
Specific Examples: Test with various text prompts to create diverse visuals.
Parameter Adjustments: Experiment with different settings to refine results.
Creative Applications: Use the model for unique and innovative visual projects.
Integration Scenarios: Combine with other tools for enhanced workflows.
Key considerations
Content Generation: The model is capable of generating both SFW and NSFW content. Users should exercise discretion and comply with relevant guidelines and regulations when generating content.
Limitations
Image Quality:
The model performs best with high-resolution and well-structured input images. Low-quality or blurry inputs may result in suboptimal outputs or artifacts.
Prompt Specificity:
Realistic Vision heavily relies on detailed and well-crafted prompts. Ambiguous or overly general prompts may produce inconsistent results.

Output Format: PNG