Nano Banana 2 & Pro Sale — 15% OFF | Apr 1–15 Only
Se connecter
Accueil/Explorer/Wan 2.7 Models/alibaba/wan-2.7/text-to-image
text-to-image

text-to-image

Alibaba WAN 2.7

alibaba/wan-2.7/text-to-image

Alibaba WAN 2.7 Text-to-Image generates high-quality images from text prompts with thinking mode for enhanced image quality. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.

Input
width
height
2048 × 2048 px
Range: 512 - 4096
Enable thinking mode for enhanced reasoning and better image quality. Increases generation time.

Idle

Close-up portrait of a model whose face is partially covered in flowing liquid metal or an iridescent, second-skin-like substance. She has otherworldly, light purple eyes and stares directly into the camera. The background is completely blurred out, leaving only a soft halo of light. The lighting is even and ethereal, as if from a bioluminescent source. Inspired by the style of Nick Knight, the image emphasizes surreal textures and subtle color gradients, exceptionally sharp, with breathtaking detail, 16K.\n

Votre requête coûtera $0.03 par exécution.

Pour $1 vous pouvez exécuter ce modèle environ 33 fois.

Encore une chose :

ExemplesTout voir

a group of animals standing in line to buy coffee, side view, anthropomorphic animals, a dog, a cat, a raccoon and a rabbit waiting in a queue, holding coffee cups, modern coffee shop counter, barista in background, casual daily scene, natural behavior, soft morning light, realistic environment, cinematic composition, 35mm photography, shallow depth of field, warm tones, high detail, ultra realistic
Close-up portrait of a model whose face is partially covered in flowing liquid metal or an iridescent, second-skin-like substance. She has otherworldly, light purple eyes and stares directly into the camera. The background is completely blurred out, leaving only a soft halo of light. The lighting is even and ethereal, as if from a bioluminescent source. Inspired by the style of Nick Knight, the image emphasizes surreal textures and subtle color gradients, exceptionally sharp, with breathtaking detail, 16K.\n
A mix collage with rapper, diamond, concert, neons, scratch paper, lyrics on paper, racing cars, money, and girls with a futuristic vibe
A fair-skinned model with classical beauty, lounging on a velvet chaise lounge, surrounded by old books and withered roses. She is wearing a baroque-style lace gown, her expression is languid and contemplative. The scene is a dim, old library, with a single stream of Rembrandt-style light from a side window illuminating her face and figure. Composition inspired by a John William Waterhouse painting, rich in narrative. The overall tones are deep and heavy, with strong chiaroscuro, creating an oil painting texture and detail.

README

Wan 2.7 Text-to-Image

Wan 2.7 Text-to-Image is Alibaba's advanced text-to-image generation model, producing high-quality, detailed images from natural language descriptions. With custom size control, built-in thinking mode, and support for a wide range of aspect ratios, it covers everything from social media content to high-resolution creative assets.

Why Choose This?

  • High-quality image generation Produces richly detailed, visually coherent images with accurate composition, lighting, and texture from text descriptions.

  • Thinking mode for smarter generation Built-in thinking mode enables the model to reason about prompt intent before generating, producing more coherent compositions and better prompt adherence.

  • Custom size output Set output width and height directly (512–4096 per dimension) to match any format — banners, thumbnails, portraits, or widescreen compositions.

  • Broad aspect ratio support Presets include 1:1, 16:9, 9:16, 4:3, 3:4, 3:2, and 2:3 for any platform or delivery format.

  • Seeded iteration Use a fixed seed to refine style and layout with more repeatable variations.

  • Prompt Enhancer Built-in tool to automatically improve your text descriptions for richer results.

Parameters

ParameterRequiredDescription
promptYesText description of the image subject, scene, style, lighting, and mood.
sizeNoOutput dimensions (width × height). Range: 512–4096 per dimension. Default: 1024×1024.
thinking_modeNoEnable thinking mode for enhanced reasoning and better image quality. Default: enabled.
seedNoFixed seed for repeatable iterations. Use -1 for a random seed.

How to Use

  1. Write your prompt — describe the subject, setting, and style. Use the Prompt Enhancer for better results.
  2. Choose a size — select a preset aspect ratio or set custom width and height to match your target format.
  3. Set thinking_mode — leave enabled (default) for best quality, or disable for faster generation.
  4. Set seed (optional) — fix a seed to make iterative prompt refinements more comparable.
  5. Submit — review the result and iterate as needed.

Pricing

Just $0.03 per generated image.

Best Use Cases

  • Social Media Content — Create platform-optimized visuals across multiple aspect ratios in one workflow.
  • Marketing & Advertising — Produce on-brand campaign visuals quickly without a photoshoot.
  • Concept Art & Storyboarding — Rapidly visualize scenes, characters, and environments from text descriptions.
  • E-commerce — Generate product lifestyle imagery and scene compositions for storefronts.
  • Creative Exploration — Rapidly prototype visual ideas and styles from detailed prompts.

Pro Tips

  • Structure your prompt as subject + environment + style: "A modern tea shop interior, warm afternoon light, minimalist wood design, cinematic photography."
  • Add camera and composition cues when framing matters: "wide shot, shallow depth of field, 35mm film look."
  • Keep thinking_mode enabled for best results — disable it only if generation speed is the priority.
  • Fix a seed while tweaking your prompt to isolate the effect of each change.
  • Generate multiple variations at smaller sizes to explore compositions before committing to a final render.

Notes

  • Only prompt is required; all other parameters are optional.
  • Output size range is 512–4096 pixels per dimension, with total pixels between 768×768 and 2048×2048 and aspect ratio between 1:8 and 8:1.
  • Thinking mode is enabled by default and improves quality but adds some latency.

Related Models