Lumiet is an AI image generation tool centered on chat-based interaction, with the product name AI Image Generator. Its core feature is that no prompt engineering is required. Users describe the image they want in everyday language, the AI returns an image within seconds, and they can keep refining it through continued conversation until they are satisfied and download it.
Uses and Target Users
Lumiet is mainly aimed at e-commerce sellers, content creators, and marketers. According to the official description, it can be used for scenarios such as e-commerce product photography, content creator thumbnails, social media marketing, and educational illustrations.
Main Capabilities
- Conversational image generation: Users enter a description, the AI generates an image, and then they can request changes in natural language (such as "make the lighting warmer" or "add a bit of shadow"). The AI remembers the context of the entire session.
- Multiple model options: Four models are built in and can be switched within the same session.
- Flux Fast: Only 1 credit per image, ultra-low cost.
- Nano Banana: Fast drafts, generates 1K images in 2-5 seconds, 5 credits per image.
- Nano Banana 2: Professional quality at flash speed, generates in 4-6 seconds, supports 1K-4K, 7-14 credits per image.
- Nano Banana Pro: Adds a deep reasoning stage, supports up to 4K (3840x2160), excels at complex text rendering, 10-20 credits per image.
- High-resolution output: Supports downloads at up to 4K resolution.
- Text rendering: Nano Banana Pro can accurately render text such as logos, titles, and multilingual labels.
- Character consistency: After creating a character once, facial and style consistency can be maintained across different scenes.
- Upload reference images for editing: Existing photos can be uploaded to modify the background, lighting, or style through conversation, and the AI preserves the parts the user did not mention. The maximum number of reference images is up to 4 for Nano Banana, and up to 14 for Nano Banana 2 and Nano Banana Pro.
- Multiple aspect ratios: Provides 7 preset ratios, suitable for different platforms.
How to Use
Users directly describe their needs in natural language in the chat box, with no special syntax required. For example: "A product photo of a coffee cup on a white background, with morning light coming from the left." After the AI returns an image, users can continue to request changes through conversation. They can also upload a reference image and then describe what they want to change. All messages accumulate based on context, and the AI remembers preferences from the session. Generated images can be downloaded using credits, and credits are calculated by modAspect ratio and resolution consumption.
Use Cases
- E-commerce product photography: Upload product photos, test multiple backgrounds through conversation, export white-background or lifestyle images for platforms such as Amazon and Shopify.
- Content creator thumbnails: Describe a thumbnail concept, get multiple options in seconds, refine text and layout through conversation, then export high-resolution images.
- Social media marketing: Generate multiple aspect-ratio variants from one prompt, such as 1:1, 9:16, and 16:9, and lock in a brand style to mass-produce a week's worth of content.
- Educational illustrations: After describing a concept, clear labeled illustrations can be generated. Nano Banana Pro can accurately render annotations, and the same diagram can be regenerated in different language versions.
Pricing and Limitations
Lumiet uses a credit system. New accounts receive 10 free credits and do not require a credit card. Paid plans: the Pro plan starts at $9.9 per month, includes 500 credits, and can generate about 100 Nano Banana images; higher-tier plans are 2000 credits/month. Specific prices are subject to the official page. Generated images are allowed for commercial use and include an invisible digital watermark (SynthID) for AI provenance tracking, which does not affect commercial use. Other limitations such as generation speed and queue priority vary by plan. The official statement says more models and video generation features are being evaluated.
