VIDU Reference-to-Image Q2 可根据一张或多张参考图像生成高质量的新图像。它会保留参考图像的关键身份特征、结构或风格,同时创作全新场景、变体或增强构图。该模型非常适合角色一致性、对象重新诠释、风格化重设计,以及由参考Входные параметры引导的电影感重现。
About this model
Vidu Reference-to-Image Q2 是一款先进的Изображение в Изображение模型,旨在将一张或多张参考图像转换为全新的构图。借助最先进的 AI 和深度学习技术,该模型在创作增强变体、全新场景或电影感重现时,能够保留Входные параметры图像的关键身份特征、结构和风格。其底层技术可确保参考图像中的复杂细节和艺术细微差异得到保留,因此非常适合需要一致性与创意的应用场景。
除了稳健的技术基础,VIDU Reference-to-Image Q2 还具备显著的营销优势。它能够根据多个Входные параметры生成高质量且风格一致的图像,支持从娱乐内容中的角色一致性到产品设计中的对象重新诠释等多种用途。每次生成的价格为 $0.032,兼具竞争力和价值,在不牺牲Выходные данные质量的情况下为用户提供出色的效果。
Cost analysis
| Provider | Cost | Notes |
|---|---|---|
| muapiapp | $0.032 per generation | muapiapp is 20-50% more affordable than competitors while delivering comparable or superior output quality. |
| Fal.ai | $0.045 per generation | Fal.ai charges close to this price point, making muapiapp a cost-efficient alternative with similar high-quality results. |
| Replicate | $0.045 per generation | Replicate's pricing is nearly identical to Fal.ai, meaning muapiapp offers a 20-50% cost saving compared to these providers. |
muapiapp is 20-50% more affordable than competitors while delivering comparable or superior output quality.
Fal.ai charges close to this price point, making muapiapp a cost-efficient alternative with similar high-quality results.
Replicate's pricing is nearly identical to Fal.ai, meaning muapiapp offers a 20-50% cost saving compared to these providers.
** Competitor pricing is estimated based on similar model architectures and usage tiers.
Configuration schema
| Parameter | Type | Description | Default |
|---|---|---|---|
| Промпт | string | 描述Изображение内容的文本Промпт。 | Create a new scene where the masked wanderer stands inside an ancient stone observatory illuminated by rotating celestial beams; preserve the character’s clothing style and silhouette while adding glowing runes carved into the walls, mist swirling across the floor, and a dramatic cosmic light shaft from above; cinematic composition, high detail. |
| Изображение URL | array | 上传或提供参考Изображение,用于Изображение转Изображение生成。 | https://d3adwkbyhxyrtq.cloudfront.net/webassets/videomodels/vidu-q2-reference-to-image-in.jpg |
| Соотношение сторон | Enum (9 options) | Выходные данныеИзображение的Соотношение сторон。 | 1:1 |
| Разрешение | Enum (3 options) | 生成Изображение的目标Разрешение。 | 1k |
描述Изображение内容的文本Промпт。
Create a new scene where the masked wanderer stands inside an ancient stone observatory illuminated by rotating celestial beams; preserve the character’s clothing style and silhouette while adding glowing runes carved into the walls, mist swirling across the floor, and a dramatic cosmic light shaft from above; cinematic composition, high detail.上传或提供参考Изображение,用于Изображение转Изображение生成。
https://d3adwkbyhxyrtq.cloudfront.net/webassets/videomodels/vidu-q2-reference-to-image-in.jpgВыходные данныеИзображение的Соотношение сторон。
1:1生成Изображение的目标Разрешение。
1kDeveloper documentation
准备Входные параметры
配置技术设置
开始生成
vidu-q2-reference-to-image)提交Входные параметры。查看并解读结果
Frequently asked
模型会使用先进的深度学习技术捕捉并保留参考Входные параметры的结构和风格等关键特征,确保生成图像体现原始图像的核心身份。
详细且具有描述性的文本Промпт能够带来最佳效果。加入关于风格、构图和期望变体的具体指导,有助于模型让最终Выходные данные更符合你的构想。
有。你最多可以提供 7 张参考图像,从而组合多个灵感来源,同时确保模型能够高效处理这些Входные параметры。
你可以选择 1k、2k 或 4k 分辨率。具体选择取决于项目的质量要求和Выходные данные媒介。