Alt Text Generator MCP
Generate accessible alt text and rich descriptions for single images using local Qwen2.5-VL vision AI. Returns JSON with alt_text (concise, screen-reader friendly) and description (detailed narrative). No external API calls — fully local.
Parameters
| Parameter | Type | Required | Description |
|---|---|---|---|
image |
string | ✅ | Image URL or base64-encoded data |
context |
string | ❌ | Surrounding page context for relevance tuning |
max_length |
integer | ❌ | Max chars for alt_text (default: 125) |
Example Response
{
"alt_text": "Engineer reviewing circuit board under magnification lamp",
"description": "A female engineer in a clean room environment examines a green PCB using a magnification lamp. She wears an ESD wrist strap and blue nitrile gloves. Soldering tools are visible in the background.",
"tags": ["engineering", "electronics", "pcb", "clean-room"],
"confidence": 0.97
}
Use Cases
- Automated image accessibility for blog CMS plugins
- Legal document image captioning for court filings
- Medical image labeling for EMR systems (on-premise safe)
Access
# x402 endpoint — pay $0.01 USDC per call (Base mainnet)
POST https://x402.ntriq.co.kr/alt-text
# Service catalog
curl https://x402.ntriq.co.kr/services
x402 micropayments — USDC on Base, gasless EIP-3009