AI Generation
Chat, image generation, video generation, and history
Overview
The product includes three AI generation modes: chat, image generation, and video generation. Paid generation uses credits and updates the result status so users can review completed and failed work later.
Chat
Chat is useful for drafting text, exploring ideas, running text tasks, and checking prompt behavior quickly.
| Behavior | Description |
|---|---|
| New chat | Users can start a new AI conversation directly |
| Follow-up messages | A conversation can keep context across messages |
| Credit cost | Each user message costs 10 credits |
| Failed result | Failed generations are not shown as successful answers |
Image Generation
Image generation currently focuses on reference-image editing. Users provide a prompt and a reference image, then the product adjusts style, content, or visual direction based on the prompt.
| Item | Description |
|---|---|
| Input | Prompt and reference image |
| Options | Size, watermark, and related generation settings |
| Credit cost | Each image generation costs 20 credits |
| Result | Successful images appear in generation history |
Video Generation
Video generation supports both text-to-video and reference-image-to-video workflows. After submission, the task moves into processing and users can wait on the page or check the result later.
| Item | Description |
|---|---|
| Text-to-video | Generate a video from a prompt only |
| Image-to-video | Use a reference image to guide the subject and framing |
| Credit cost | Each video generation costs 50 credits |
| Status | Common states include pending, processing, completed, and failed |
Generation History
Image and video results are saved in generation history. A history record usually includes the generation type, prompt, reference image, result link, status, credits used, and generation settings.
Users can revisit results, download media, or reuse prompts from history. Failed records keep their status so users can decide whether to try again.