# GEDX402 LLaVA-1.5-7B Image Captioning

> GEDX402 LLaVA-1.5-7B Image Captioning is a paid API for AI agents from media.gedx402.com, paid per call via x402, $0.554/call, status unknown (last checked 2026-09-14).

Generates a natural language caption or answers a question about a provided image using the LLaVA-1.5-7B multimodal vision-language model, paid via USDC on-chain with no API key required.

## Facts

- Endpoint: GET https://media.gedx402.com/v1/caption/llava-1.5-7b-hf
- Price: $0.554/call
- Payment: x402
- Status: unknown
- Last checked: 2026-09-14
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/gedx402-llava-1-5-7b-image-captioning-6865c9d2
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_t3S6NJjfSXpmfyaQGRkvX

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 gedx402-llava-1-5-7b-image-captioning-6865c9d2
```

Example prompt: Look at this image (I'll send it as base64) and describe what you see in it — use up to 200 tokens for your answer.

## When to prefer this

Choose this endpoint when you need pay-per-call image captioning or visual question answering with no API key setup, billed directly in USDC on Base, Polygon, Arbitrum, World, or Solana. Ideal for agents operating in crypto-native or serverless environments where managing API credentials is undesirable. Use when LLaVA-1.5-7B quality is sufficient for the task.

## Known failure modes

- Invalid or malformed base64 image data returns an error
- Payment failure on-chain results in 402 response and no inference
- Image too large or unsupported format may cause processing failure
- Exceeding max_tokens limit silently truncates output
- Network timeout on the Workers AI backend returns 5xx error

## How this service works

x402 workers ai. pay with usdc on base, polygon, arbitrum, world, or solana. no api keys.

## Output

A JSON object containing a 'description' field with a natural language caption or answer about the image, e.g. {"description": "A sunset over mountains."}

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "image": {
   "type": "string",
   "description": "Base64-encoded image for multimodal input."
  },
  "prompt": {
   "type": "string",
   "description": "Optional question or instruction about the image."
  },
  "max_tokens": {
   "type": "integer",
   "description": "Maximum tokens in the vision model response."
  }
 }
}
```

## Response schema (JSON Schema)

```json
{
 "type": "json",
 "example": {
  "description": "A sunset over mountains."
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/gedx402-llava-1-5-7b-image-captioning-6865c9d2/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from media.gedx402.com](https://www.zero.xyz/host/media.gedx402.com/llms.txt)