# MiniChan Code-from-Image — NVIDIA NIM MiniMax-M3

> MiniChan Code-from-Image — NVIDIA NIM MiniMax-M3 is a paid API for AI agents from app-minichi.vercel.app, paid per call via x402, $0.2/call, status unknown (last checked 2026-10-13).

Generates working source code from a screenshot or diagram image using a 428B MoE multimodal model

## Facts

- Endpoint: POST https://app-minichi.vercel.app/api/code-from-image?utm_source=zero.xyz
- Price: $0.2/call
- Payment: x402
- Status: unknown
- Last checked: 2026-10-13
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/minichan-code-from-image-nvidia-nim-minimax-m3-ddbe3e9e
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_pVVNk9K8qxmzf9efykEn0

Status and success rate cover calls made through Zero and Zero's own probes. Third-party monitors may report differently.

## How to call it through Zero

Zero handles the 402 payment challenge and records the run. With the Zero CLI installed (`npm i -g @zeroxyz/cli`):

```sh
zero fetch --capability minichan-code-from-image-nvidia-nim-minimax-m3-ddbe3e9e -d '<json body>'
```

Example prompt: Here's a screenshot of a login form UI — can you generate the HTML and CSS code to recreate it exactly?

## When to prefer this

Choose this endpoint when you have an image (screenshot, photo, mockup, or diagram) and need to produce executable or usable source code from it. It is ideal for reverse-engineering UI designs into frontend code, transcribing handwritten or photographed code, or implementing logic from whiteboard diagrams. Prefer this over text-based code generation when the primary input is visual rather than a written specification.

## Known failure modes

- Image URL is inaccessible or returns a non-image response — model cannot process it
- Image is too low resolution or blurry — output code may be inaccurate or incomplete
- Image contains no recognizable code or UI structure — model may return a best-guess or error
- Payment of 0.2 USDC via x402 on Base fails — request is rejected before processing
- Ambiguous diagram with no clear programming context — generated code may not match intent

## How this service works

NVIDIA NIM MiniMax-M3 (428B MoE) powered API for AI agents. 15 multimodal endpoints — text, vision, code, reasoning, creative. Pay with USDC on Base via x402.

## Output

Returns generated source code reconstructed or derived from the input image, such as HTML/CSS for UI screenshots, Python for diagrams, or verbatim code extracted from photos of code snippets.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "required": [
  "image_url"
 ],
 "properties": {
  "image_url": {
   "type": "string",
   "description": "URL of the image to analyze"
  }
 }
}
```

## Response schema (JSON Schema)

```json
{
 "type": "object"
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/minichan-code-from-image-nvidia-nim-minimax-m3-ddbe3e9e/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from app-minichi.vercel.app](https://www.zero.xyz/host/app-minichi.vercel.app/llms.txt)