Struggling with lackluster results from your text to image generator? If you’re passionate about AI image creation, you know a great prompt is key to stunning visuals. PromptEnhancer, an open-source tool from Tencent Hunyuan, transforms vague prompts into clear, professional instructions, enhancing outputs for tools like Stable Diffusion, DALL-E, or Midjourney.
This guide explores how PromptEnhancer revolutionizes text to image workflows, covering its features, setup, and usage tips. Perfect for beginners or pros searching for “text to image AI tools” or “AI art prompt optimization,” let’s dive in!
What is PromptEnhancer?
PromptEnhancer refines prompts for text to image generators using chain-of-thought reasoning. It restructures inputs while preserving intent, delivering clearer prompts for better AI-generated images. Supporting English and Chinese, it’s ideal for global users seeking “best AI text to image prompts” or “AI image generation tips.”
Key Features
Why choose PromptEnhancer for “text to image generator enhancements”? Here’s what makes it stand out:
- Dual-Mode Support: Enhances text-to-image and image-to-image editing prompts.
- Intent Preservation: Keeps subjects, actions, styles, and attributes intact.
- Robust Parsing: Ensures reliable outputs for complex prompts.
- Flexible Deployment: Offers full-precision (7B/32B) and quantized GGUF models for lower memory use.
- Memory Efficiency: GGUF models cut VRAM usage by 50-75%, ideal for RTX 3090/4090 GPUs.
Installation
Set up PromptEnhancer easily:
Standard Installation
pip install -r requirements.txtGGUF Installation (Memory-Efficient)
chmod +x script/install_gguf.sh && ./script/install_gguf.shModel Selection
Choose the right model for your needs, based on searches like “best AI text to image models”:
| Model | Size | Quality | Memory | Best For |
|---|---|---|---|---|
| PromptEnhancer-7B | 13GB | High | 8GB+ | Balanced performance |
| PromptEnhancer-32B | 64GB | Highest | 32GB+ | Research, top quality |
| 32B-Q6_K (GGUF) | 27GB | Excellent | 27GB+ | RTX 4090, pros |
| 32B-Q4_K_M (GGUF) | 20GB | Good | 20GB+ | RTX 3090, budget setups |
Download the 7B model:
huggingface-cli download tencent/HunyuanImage-2.1/reprompt --local-dir ./models/promptenhancer-7bFor image-to-image editing:
huggingface-cli download PromptEnhancer/PromptEnhancer-Img2img-Edit --local-dir ./models/promptenhancer-img2img-editHow to Use PromptEnhancer
Text-to-Image
from inference.prompt_enhancer import HunyuanPromptEnhancer
enhancer = HunyuanPromptEnhancer(models_root_path="./models/promptenhancer-7b", device_map="auto")
user_prompt = "A race car speeding on a city track"
new_prompt = enhancer.predict(prompt_cot=user_prompt, temperature=0.7, top_p=0.9, max_new_tokens=256)
print("Enhanced:", new_prompt)Image-to-Image Editing
from inference.prompt_enhancer_img2img import PromptEnhancerImg2Img
enhancer = PromptEnhancerImg2Img(model_path="./models/promptenhancer-img2img-edit", device_map="auto")
edit_instruction = "Remove the watermark"
image_path = "./examples/sample_image.png"
enhanced_prompt = enhancer.predict(edit_instruction=edit_instruction, image_path=image_path, temperature=0.1, top_p=0.9, max_new_tokens=2048)
print("Enhanced:", enhanced_prompt)GGUF (Quantized)
from inference.prompt_enhancer_gguf import PromptEnhancerGGUF
enhancer = PromptEnhancerGGUF(model_path="./models/PromptEnhancer-32B.Q6_K.gguf", n_ctx=1024, n_gpu_layers=-1)
enhanced_prompt = enhancer.predict("woman in jungle", temperature=0.3, top_p=0.9, max_new_tokens=512)
print("Enhanced:", enhanced_prompt)Why Use PromptEnhancer?
PromptEnhancer delivers high-quality AI art by refining prompts for niches like “realistic photo portraits” or “concept art prompts.” Its memory-efficient GGUF models make it accessible for standard hardware, solving issues for those searching “improve AI generated images.”
Final Thoughts
PromptEnhancer turns basic prompts into professional-grade instructions, enhancing your text to image generator results. Download it from GitHub and elevate your AI art today. Share your favorite “text to image prompts” in the comments, and let’s explore the future of AI image creation together!
