llm
Advanced Prompt Optimizer for image and video generation. Automatically enhances prompts for clarity, structure, composition, motion dynamics, and style control—producing significantly better outputs across models like FLUX, Wan, Kling, Veo, Seedance, and more. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
Input
Idle
A glamorous, confident woman with voluminous blonde curls strides down a Parisian street at dusk, exuding 1950s cinematic allure. She wears a sleek black off-the-shoulder satin gown, paired with striking red high heels and holds a sparkling silver clutch. Smoke curls from her cigarette as she gazes forward, embodying bold fashion and urban sophistication. Behind her, the neon-lit “BALAZLE” theater sign glows against the twilight sky, while a vintage blue car idles nearby. The scene is captured in a cinematic, high-contrast style with dynamic camera movement, emphasizing motion and atmosphere.
您的请求将花费 $0.001 每次运行。
使用 $1 您可以运行此模型大约 1000 次。
示例查看全部
README
Image & Video Prompt Optimizer API
Overview
The WaveSpeedAI Prompt Optimizer enhances prompts specifically for image and video generation workflows. It restructures and enriches your input prompt to improve visual clarity, composition, cinematic framing, lighting, camera movement, and style consistency.
Designed for models such as Nano Banana, Seedream, FLUX, Wan, Kling, Veo, Seedance, Hailuo, and other leading visual-generation systems.
Key Features
- Improves visual composition (lighting, framing, perspective)
- Expands prompts with cinematic and artistic attributes
- Enhances style, mood, and motion for video models
- Reduces ambiguity and ensures consistent reproducible outputs
- Works seamlessly in automated pipelines and API workflows
Why Use It?
Raw user prompts often produce unstable or low-quality visual outputs. The Prompt Optimizer standardizes and enriches prompts, leading to:
- Better image detail and fidelity
- More stable, coherent video motion
- Consistent art direction across batches
- Improved adherence to user intent
Usage
Send any raw text prompt to the /v1/prompt/optimize endpoint and receive a fully optimized version ready for visual generation models.