# Image Analysis – Caption, Tags & Object Detection

> Image Analysis – Caption, Tags & Object Detection is a paid API for AI agents from manyfish-seller.redcliff-d7368db1.westus2.azurecontainerapps.io, paid per call via x402, $0.004/call, status unknown (last checked 2026-09-14).

Analyzes a base64-encoded JPEG or PNG image and returns a descriptive caption, visual tags, and detected objects with confidence scores.

## Facts

- Endpoint: POST https://manyfish-seller.redcliff-d7368db1.westus2.azurecontainerapps.io/api/v1/image/analyze
- Price: $0.004/call
- Payment: x402
- Status: unknown
- Last checked: 2026-09-14
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/image-analysis-caption-tags-object-detection-00d7e19e
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_hBL8AEkL2HtOOeAeMagHQ

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 image-analysis-caption-tags-object-detection-00d7e19e -d '<json body>'
```

Example prompt: Analyze this JPEG image for me — give me a caption describing the scene, any relevant visual tags, and a list of objects you can detect, only including results with at least 70% confidence.

## When to prefer this

Choose this endpoint when you need a combined response of caption, tags, AND object locations in a single call, rather than separate endpoints for each task. It is particularly useful for enriching image metadata, powering search indexing, or feeding downstream NLP pipelines. Prefer the dedicated object-detection endpoint if you only need bounding boxes, or the caption-only endpoint if you need a single sentence description.

## Known failure modes

- Invalid base64 string — malformed or non-base64 payload returns a 400 error
- Unsupported content type — only image/jpeg and image/png are accepted
- Image too large — base64 payload exceeding ~10MB may be rejected
- Low-quality or ambiguous images may return sparse tags or a generic caption
- Confidence threshold set too high may return empty object lists

## How this service works

Return a caption, visual tags, and located objects for a supplied image.

## Output

Returns a natural-language caption describing the image, a list of visual tags (keywords describing scene elements, colors, styles, etc.), and a list of detected objects each with a label and confidence score. Results are filtered by the supplied minimum confidence threshold (default 0.5).

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "contentType": {
   "enum": [
    "image/jpeg",
    "image/png"
   ]
  },
  "imageBase64": {
   "type": "string",
   "pattern": "^(?:[A-Za-z0-9+/]{4})*(?:[A-Za-z0-9+/]{2}==|[A-Za-z0-9+/]{3}=)?$",
   "maxLength": 13981016,
   "minLength": 8
  },
  "minimumConfidence": {
   "type": "number",
   "default": 0.5,
   "maximum": 1,
   "minimum": 0
  }
 }
}
```

## Response schema (JSON Schema)

```json
{
 "type": "json",
 "example": {
  "tags": [
   {
    "name": "text",
    "confidence": 0.97
   }
  ],
  "image": {
   "width": 200,
   "height": 120,
   "contentType": "image/png"
  },
  "caption": {
   "text": "A red circle and the word hello on a white background.",
   "confidence": 0.92
  },
  "objects": [],
  "language": "en"
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/image-analysis-caption-tags-object-detection-00d7e19e/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from manyfish-seller.redcliff-d7368db1.westus2.azurecontainerapps.io](https://www.zero.xyz/host/manyfish-seller.redcliff-d7368db1.westus2.azurecontainerapps.io/llms.txt)