Async AI image generation (text-to-image and image-to-image). Submit a job to get a task_id, then poll status to get an OSS download URL.
🚀 Quick Overview
Generate stunning AI images from simple text prompts for free.
Powered by Gemini-level AI image generation technology.
Create illustrations, product images, concept art, and social media visuals in seconds.
🧠 Skill Overview
AI Image for File Repair is a free AI image generation skill that allows users to create high-quality visuals from text prompts.
Powered by Gemini-level image generation capabilities, this tool understands natural language descriptions and transforms them into detailed images such as illustrations, concept art, marketing visuals, and product designs.
Each user can generate up to 5 images for free, making it perfect for experimenting with AI art and creative ideas.
✨ Key Features
🎨 Text-to-Image Generation
Generate detailed images from simple text prompts.
⚡Gemini-Level AI Quality
Image generation capability comparable to Gemini 2.5.
🆓 Free AI Image Generation
Generate up to 5 images for free.
🎭Multiple Styles
Supports various styles including:
🎨 Example Prompts
Try prompts like:
🚀 Use Cases
This skill is perfect for:
💡 Why Use This Skill
⚡ Quick Start
1️⃣ Enter a detailed prompt
2️⃣ Generate your AI image
3️⃣ Download or use the generated image
🔥 Trending Prompts
🏷 Tags
ai-lmage-for-file-repair exposes an async, job-based image generation API:
task_idresult.file_url (OSS URL) for download/displayerror payloadAI_IMAGE_HOST environment variable, e.g. http://localhost:9000x-api-key with the value of AI_IMAGE_API_KEYHeader example:
-H "x-api-key: ${AI_IMAGE_API_KEY}"
Configuration example (local development):
AI_IMAGE_HOST=https://file-repiar-wfaxciicwy.ap-southeast-1.fcapp.run(this is the default when you run this repo'sindex.jslocally)AI_IMAGE_API_KEY=s8cxRtUxNcNEL4um
x-forwarded-for; if absent it falls back to req.ipx-forwarded-for is correctly forwarded, otherwise multiple users may share the same outbound IP quotaGET /quota: check today's usage and remaining quotaPOST /generate: submit a generation job (text-to-image / image-to-image)GET /status/:taskId: check job status and retrieve the download URLGET /quotaResponse:
{
"client_id": "ip:203.0.113.10",
"date": "2026-03-10",
"used": 1,
"limit": 3,
"remaining": 2
}
POST /generateapplication/jsonRequest body (text-to-image):
"1:1", "16:9"For image-to-image, add:
data:image/...;base64, prefix)input_image_base64 is provided, string): e.g. image/png, image/jpegSuccess response:
{ "task_id": "xxxxxxxxxxxxxxxx" }
GET /status/:taskIdCommon fields:
queued / running / succeeded / failednull on failurenull on successSuccess example:
{
"task_id": "xxxxxxxxxxxxxxxx",
"status": "succeeded",
"created_at": 1760000000000,
"updated_at": 1760000005000,
"started_at": 1760000001000,
"finished_at": 1760000004500,
"error": null,
"result": {
"file_url": "https://ts-xxx.oss-cn-shenzhen.aliyuncs.com/ai-photo/20260117070932_user-card-bg.png",
"file_name": "user-card-bg.png",
"mime_type": "image/png"
}
}
Failure example (shape):
{
"task_id": "xxxxxxxxxxxxxxxx",
"status": "failed",
"error": {
"message": "Upstream HTTP 401",
"code": "UPSTREAM_ERROR",
"statusCode": 401,
"upstream": {}
},
"result": null
}
GET /quota and ensure remaining > 0POST /generate to get a task_idGET /status/:taskId
queued/running: poll every 1–2 seconds (recommend polling up to 30–60 seconds; you can retry later)succeeded: download/display via result.file_urlfailed: show error.message and stopcurl -X POST "${AI_IMAGE_HOST}/generate" \
-H "Content-Type: application/json" \
-H "x-api-key: ${AI_IMAGE_API_KEY}" \
-d '{
"prompt": "一只戴墨镜的柯基，赛博朋克霓虹灯风格，高清",
"aspect_ratio": "1:1"
}'
$body = @{
prompt = "A corgi wearing sunglasses, cyberpunk neon, high detail"
aspect_ratio = "1:1"
} | ConvertTo-Json -Depth 10
Invoke-RestMethod -Method Post "$env:AI_IMAGE_HOST/generate" `
-Headers @{ "x-api-key"=$env:AI_IMAGE_API_KEY } `
-ContentType "application/json" `
-Body $body
$b64 = [Convert]::ToBase64String([IO.File]::ReadAllBytes("input.png"))
$body = @{
prompt = "Enhance and restore this image, keep the subject, photo-realistic"
input_image_base64 = $b64
input_image_mime_type = "image/png"
} | ConvertTo-Json -Depth 10
Invoke-RestMethod -Method Post "$env:AI_IMAGE_HOST/generate" `
-Headers @{ "x-api-key"=$env:AI_IMAGE_API_KEY } `
-ContentType "application/json" `
-Body $body
Tip: the
base64flags differ across platforms. The PowerShell example above is the most portable for Windows.
IMG_B64="$(base64 -i input.png | tr -d '\n')"
curl -X POST "${AI_IMAGE_HOST}/generate" \
-H "Content-Type: application/json" \
-H "x-api-key: ${AI_IMAGE_API_KEY}" \
-d "{
\"prompt\": \"Enhance and restore this image, keep the subject, photo-realistic\",
\"input_image_base64\": \"${IMG_B64}\",
\"input_image_mime_type\": \"image/png\"
}"
curl -L "<file_url>" -o "output.png"
prompt; or missing input_image_mime_type for image-to-image)x-api-keyerror.upstream and server logs)If you require more extensive or advanced format repair support, or you need audio/image repair, please visit: https://bit.ly/4roS6Rv
We offer the following services:
●Effortlessly repair videos, photos and files that are blurry, overexposed and corrupted. ●AI enhance photo quality, sharpen detail, and colorize photos. ●Effectively improve video quality and boost resolution up to 8K. ●Extend, fill, and restyle your images with AI. No quality loss. ●Expertly repair corrupted files, including Word, Excel, PPTX, EPUB, PDF, PSD, PSB, and more. ●Seamlessly repair corrupted audio files in M4A, MP3, AAC, WAV and FLAC formats. ●Convert PNG, HEIC, CR2, BMP -> JPG for repair & enhance
ZIP package — ready to use