WaveSpeed.ai
DashboardExploreInspiration
Home/Explore/Qwen AI Models/wavespeed-ai/jib-mix-qwen-image/text-to-image-lora

text-to-image

wavespeed-ai/jib-mix-qwen-image/text-to-image-lora

Jib Mix Qwen LoRa — more natural pretty faces (Much better at Asian faces) model for next-gen text-to-image generation with LoRA.

width
height
If set to true, the function will wait for the result to be generated and uploaded before returning the response. It allows you to get the result directly in the response. This property is only available through the API.
If enabled, the output will be encoded into a BASE64 string instead of a URL. This property is only available through the API.

Idle

"Image Type": "Illustration", Gufeng (Ancient Chinese style), hand-drawn, ink wash diffusion, high quality, delicate texture, diffuse gradient, storytelling, diffuse reflection, storytelling, stream of consciousness, rice paper texture, embellished with light gauze and light spots, layered, flowing ultra-long hair, gorgeous accessories and intricate complex clothing textures, peony flower demon (yaoguai), beautiful woman, exotic, extremely beautiful, beautiful and bewitching, messy hair strands, long hair, Liaozhai (Strange Tales from a Chinese Studio), Chinese supernatural gothic, dark, entwining, dark, wicked smile, horror, (Green lotus and water often meet mountains and rivers, orioles sing and swallows talk, deep feelings connect you and I also), exquisite details, deep blue color palette, accented with yellow, red dots, and green, extreme details, hand-drawn illustration, flat illustration, stream of consciousness, hazy, psychedelic, lonely world, close-up, color graffiti, particle texture, x showcase story and style, image outline extraction, dreamy style, soft light, diffuse gradient, faint light in the dark, layered, gouache, depth and dreamlike effect, obvious frosted grain texture, heavy bold color block brushstrokes, vivid colors, high contrast (chiaroscuro), glowing halo, elegant/high-class feel, high-angle view, overhead shot, visual impact, dynamic, stunning, HD details.

Your request will cost $0.025 per run.

For $1 you can run this model approximately 40 times.

One more thing:

ExamplesView all

"Image Type": "Illustration", Gufeng (Ancient Chinese style), hand-drawn, ink wash diffusion, high quality, delicate texture, diffuse gradient, storytelling, diffuse reflection, storytelling, stream of consciousness, rice paper texture, embellished with light gauze and light spots, layered, flowing ultra-long hair, gorgeous accessories and intricate complex clothing textures, peony flower demon (yaoguai), beautiful woman, exotic, extremely beautiful, beautiful and bewitching, messy hair strands, long hair, Liaozhai (Strange Tales from a Chinese Studio), Chinese supernatural gothic, dark, entwining, dark, wicked smile, horror, (Green lotus and water often meet mountains and rivers, orioles sing and swallows talk, deep feelings connect you and I also), exquisite details, deep blue color palette, accented with yellow, red dots, and green, extreme details, hand-drawn illustration, flat illustration, stream of consciousness, hazy, psychedelic, lonely world, close-up, color graffiti, particle texture, x showcase story and style, image outline extraction, dreamy style, soft light, diffuse gradient, faint light in the dark, layered, gouache, depth and dreamlike effect, obvious frosted grain texture, heavy bold color block brushstrokes, vivid colors, high contrast (chiaroscuro), glowing halo, elegant/high-class feel, high-angle view, overhead shot, visual impact, dynamic, stunning, HD details.
Adopting mechanical realism style, Gundam mecha, white primary color movie texture, red secondary color mecha body, mechanical armor structure, ultra-realistic metal material, metallic luster, showing its real and complex internal structure, cyberpunk style, future technology, mechanical armor, real movie texture, side light enhanced three-dimensional texture, 3D rendering ultra-high-definition 8K, 32k resolution parameters, sharp depth of field, movie-level noise processing,
realism, A whimsical watercolor portrait of a young woman with flowers woven into her messy blonde hair. Soft pastel colors, translucent washes, delicate lines, dreamy and ethereal expression, looking slightly upwards, white background, in the style of Alphonse Mucha.
Medium-shot portrait of a smiling woman with dark brown hair tied in a messy bun, wearing a stylish teal blue bikini and sunglasses. She is lounging on a sun chair by a crystal-clear swimming pool. Bright, direct sunlight, sharp focus, vibrant colors, luxury resort background, photorealistic.
Gritty, realistic portrait of a Viking warrior chief after a battle. Battle-worn face, scars, braided beard with metal rings. Fur-lined cloak, staring intensely forward. Cold, overcast lighting, snowy fjord background. Cinematic, style of "The Northman".
Surrealist portrait of a person's silhouette created with a double exposure effect. The inside of the silhouette is filled with a swirling nebula and galaxy of stars. Face is in profile, calm, eyes closed. Deep indigo, magenta, and cosmic blues.
Dignified portrait of a Maasai elder, weathered face with intricate tribal face paint, wearing vibrant red Shúka (cloth) and elaborate beaded necklaces. Standing against the vast, sun-drenched savanna at sunset. National Geographic style, telephoto lens, sharp focus, respectful gaze.
An Art Deco portrait of a 1920s flapper girl. Sharp bob haircut, feathered headband, long pearl necklace. Smirking confidently at a lavish Gatsby-esque party. Geometric patterns in the background. Style of a vintage "Vogue" cover illustration.
An oil painting portrait in the style of Rembrandt. An old scholar with a long white beard, kind eyes, wearing dark robes, sitting in a dimly lit library. Strong chiaroscuro (light-dark contrast), rich textures, warm tones, masterful brushstrokes, profound expression.
Full-body concept art of an ethereal Elven Queen. Long silver hair, glowing blue eyes, intricate silver armor mixed with silk robes, standing in a magical, bioluminescent forest. Powerful and graceful stance, mystical atmosphere, fantasy, highly detailed, sharp focus.
A stunningly hyper-realistic portrait of a pensive young woman, sitting by a cafe window during golden hour. Soft, warm light illuminates half of her face, strong shadows, intricate details of her flowing hair, shallow depth of field (bokeh background), shot on 85mm f/1.4 lens, cinematic atmosphere, film grain.
An intimate, high-contrast black and white close-up portrait of an elderly man with deep wrinkles. His eyes are filled with wisdom and melancholy. Dramatic side lighting (chiaroscuro), capturing the texture of his skin, style of street photography, grainy, powerful and emotional.
Medium shot portrait of a man in a black leather jacket, standing on a neon-lit Tokyo street at night. Rain-slicked pavement, vibrant neon signs reflected in his sunglasses, serious expression, looking directly at the camera, Blade Runner aesthetic, high contrast, sharp focus.

README

Jib-Mix-Qwen-Image-LoRA (Text-to-Image)

Jib-Mix-Qwen-Image-LoRA is a creative powerhouse that fuses the Qwen-Image 20B MMDiT backbone with LoRA adaptability and Jib-Mix fine-tuning. It excels at producing stunning human portraits, cinematic lighting, and expressive stylistic control—all with simple text prompts. Perfect for creators seeking world-class face generation and full artistic flexibility.

Why it looks great

  • Hybrid Jib-Mix tuning – Trained for superior facial realism, skin tone balance, and lighting fidelity; especially powerful for close-up and half-body portraits.
  • LoRA integration – Load .safetensors LoRA weights for custom characters, aesthetics, or stylizations; control blending with a simple scale slider.
  • Advanced text rendering – Exceptional bilingual performance (Chinese & English) with smooth typography inside the image.
  • Versatile aesthetics – Photorealistic, anime, painterly, or stylized—handles all with consistent detail and color harmony.
  • Emotion-aware diffusion – Captures expressions, pose subtleties, and scene atmosphere for cinematic storytelling.

Limits and Performance

  • Max resolution per job: up to 1536 × 1536 pixels
  • LoRA path: supports <owner>/<model-name> or .safetensors URL
  • LoRA scale: 0.1–1.5 (default = 1.0)
  • Output formats: JPEG / PNG / WEBP
  • Processing speed: ~6–9 seconds per image
  • Prompt: bilingual, descriptive, supports multi-line narratives

Pricing

  • $0.025 per image Each image is billed individually.

How to Use

  1. Write a prompt describing your scene or character (English or Chinese).

  2. Set width and height .

  3. Add one or more LoRAs:

    • Paste LoRA path or URL (.safetensors).
    • Adjust scale to tune blending intensity.

  4. (Optional) Set a seed for reproducibility (-1 = random).

  5. Choose output format (JPEG / PNG / WEBP).

  6. Generate → review → iterate with new LoRAs or parameters.

Pro tips for best quality

  • Use portrait-focused LoRAs to enhance realism and consistency.
  • Combine style + identity LoRAs for hybrid looks (e.g., fantasy portrait + cyberpunk mood).
  • Keep scale moderate (0.7–1.0) for natural results.
  • Fix seed to maintain identity when testing new LoRAs.

Reference

Note

  • LoRAs from Civitai or Hugging Face are supported if exported as .safetensors.
  • Model optimized for portrait, fashion, and cinematic generation—for landscapes or scenes, lower scale or combine with general LoRAs for balance.