Skip to content

Private LLM Vision · api.erb-llm.com

Privacy-first OpenAI-compatible vision

A local vision-language model on dedicated hardware understands your images - they never reach OpenAI/Anthropic or any third-party cloud. Send a public image URL plus a prompt (or OpenAI image_url messages); receive a standard chat.completion JSON answering about the image. Up to 4 images per call (more are clamped). Pay per call in USDC on Base (x402), no API key.

Ready now4 calls · 1 buyer in 30 days

Input

A public https:// image URL (simple form). Use `image`+`prompt`, or the full OpenAI `messages` array with image_url parts.
Optional output cap. Default: the full 2,048-token vision cap.
Messages
OpenAI chat messages using image_url content parts (full form). Up to 4 images per call; more are clamped to the first 4.
Optional. Default: the gateway's local vision-language model.
What to do with the image (used with `image`). Default: 'Describe this image.'

Sample response

Published by the provider. Your run returns a fresh response.

response6 keys
choices1 item
id"chatcmpl-abc123"
model"qwen/qwen3.8-27b"
object"chat.completion"
usage3 keys
completion_tokens12
prompt_tokens60
total_tokens72
x4024 keys
charged"$0.02"
inference_ms2400
tier"vision"
tokens_out12
responseobject, 6 keys
Your agent can do this on its own
Add the MCP URL in Claude Code, Cursor or Codex. Your agent can run Tools within its spending limits, or ask for approval each time. Connect your agent

Connect your agent

Add one MCP URL and your agent can find and run this tool within its spending limits. Choose “Ask each time” if you want to approve every Call.

MCP URL

https://app.aecon.ai/mcp
Terminalbash
claude mcp add --transport http --scope user agentic-economy "https://app.aecon.ai/mcp"

In Claude Code, open /mcp, select agentic-economy, then choose Authenticate.

Track record

Paid calls, last 30 days
4
Buyers, last 30 days
1
Last paid call

Reported by the x402 directory.

A$0.029 per call