# Inference402 Image Describe

> Inference402 Image Describe is a paid API for AI agents from inference402.worlddata402.workers.dev, paid per call via x402, $0.01/call, status unknown (last checked 2026-10-19).

Analyzes a public image URL and returns a natural-language description plus detected objects, powered by LLaVA vision model, paid per-call in USDC.

## Facts

- Endpoint: GET https://inference402.worlddata402.workers.dev/v1/image/describe?utm_source=zero.xyz
- Price: $0.01/call
- Payment: x402
- Status: unknown
- Last checked: 2026-10-19
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/inference402-image-describe-91ed13a8
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_rOfCsHglwjiPfXspN4INS

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 inference402-image-describe-91ed13a8
```

Example prompt: Look at this image — https://example.com/photo.jpg — and tell me what objects are in it and give me a full description of the scene.

## When to prefer this

Choose this endpoint when you need no-account, pay-per-call vision analysis of a public image URL and want structured JSON output including both an object list and a natural-language description. It is ideal for agents that handle sporadic image analysis tasks without committing to a subscription, and when payment via USDC on Base is acceptable. Prefer alternatives if you need private/uploaded image support, batch processing, or higher-resolution vision models.

## Known failure modes

- Image URL is not publicly accessible or returns a non-200 response — endpoint cannot fetch the image
- Image exceeds 5MB size limit — request rejected
- Unsupported image format (non-JPEG/PNG/WebP) — may fail or produce degraded output
- Payment not submitted in USDC on Base mainnet — 402 Payment Required returned
- Ambiguous or low-resolution image — model may return generic or inaccurate description
- Rate limit or worker timeout — transient 5xx error

## How this service works

Pay-per-call AI inference for agents: chat (Llama 3.3 70B / 3.1 8B), structured JSON, summarize, translate (incl. Egyptian Arabic), extract, classify, sentiment, keywords, rewrite, embeddings, moderation, vision and speech-to-text. USDC on Base mainnet over x402. No API key, no account.

## Output

Returns a JSON object containing the model identifier (e.g. '@cf/llava-hf/llava-1.5-7b-hf'), an array of detected object labels, and a natural-language description sentence summarizing the image contents.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "prompt": {
   "type": "string",
   "maxLength": 500,
   "description": "Optional extra instruction or question about the image"
  },
  "image_url": {
   "type": "string",
   "format": "uri",
   "description": "Public http(s) URL of a JPEG/PNG/WebP image ≤5MB"
  }
 }
}
```

## Response schema (JSON Schema)

```json
{
 "type": "json",
 "example": {
  "model": "@cf/llava-hf/llava-1.5-7b-hf",
  "objects": [
   "cat",
   "floor"
  ],
  "description": "A ginger cat sitting on a wooden floor looking at the camera."
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/inference402-image-describe-91ed13a8/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from inference402.worlddata402.workers.dev](https://www.zero.xyz/host/inference402.worlddata402.workers.dev/llms.txt)