Jetson Gateway Instructions

Use this gateway as an OpenAI-compatible chat API. Replace YOUR_API_KEY with the key provided by the gateway owner.

Connection

Base URL: https://agx1-1.taildbf607.ts.net/v1
Auth header: Authorization: Bearer YOUR_API_KEY
Chat endpoint: /chat/completions
Models endpoint: /models

Models

jetson/qwen3.5-27b
jetson/qwen3.5-35b
jetson/qwen3.5-9b-q8_0

Choose a model by setting the request body's model field to one of these ids.

Recommended Defaults

Default response limit: 8192 tokens
Maximum response limit: 16384 tokens
Default context window: 32768
Default temperature: 0.2

List Models With curl

curl https://agx1-1.taildbf607.ts.net/v1/models \
  -H "Authorization: Bearer YOUR_API_KEY"

Chat With curl

curl https://agx1-1.taildbf607.ts.net/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{
    "model": "jetson/qwen3.5-27b",
    "messages": [
      {"role": "user", "content": "hello"}
    ],
    "stream": false,
    "max_tokens": 8192,
    "extra_body": {
      "think": false,
      "num_ctx": 32768
    }
  }'

Python OpenAI SDK

from openai import OpenAI

client = OpenAI(
    api_key="YOUR_API_KEY",
    base_url="https://agx1-1.taildbf607.ts.net/v1",
)

response = client.chat.completions.create(
    model="jetson/qwen3.5-27b",
    messages=[
        {"role": "user", "content": "hello"}
    ],
    max_tokens=8192,
    temperature=0.2,
    extra_body={
        "think": False,
        "num_ctx": 32768,
    },
)

print(response.choices[0].message.content)

High-Volume Use

Use stream: true for long responses when possible.
If the gateway returns 429, wait briefly and retry with backoff.
Use x-client-name to label your application in the dashboard.

Dashboard

Open Dashboard

https://agx1-1.taildbf607.ts.net/dashboard#token=YOUR_API_KEY