ExergyNet/API Keys
API Keys

Key management

Current Key
sk-exergy-••••••••••••••••preview · full key never stored
Model Capabilities — vanguard
Loading capabilities…
Default Token Limit

Controls how many tokens the model generates per response in the Playground. 512 is a good default. Increase for longer outputs, reduce for faster responses.

512tokens
2,048 characters · max context 4 096 tokens
System Instructions

System instructions are injected before every inference call in the playground. Only one instruction can be active at a time.

Loading…
Integration — OpenAI-compatible
Python / openai SDK
from openai import OpenAI

client = OpenAI(
    api_key="sk-exergy-••••••••••••••••",
    base_url="https://dt.portal.exergynet.org/v1"
)

response = client.chat.completions.create(
    model="vanguard-engine",
    messages=[{"role": "user", "content": "..."}],
    stream=True
)

for chunk in response:
    print(chunk.choices[0].delta.content or "", end="")
cURL
curl https://dt.portal.exergynet.org/v1/chat/completions \
  -H "Authorization: Bearer sk-exergy-••••••••••••••••" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "vanguard-engine",
    "messages": [{"role": "user", "content": "Hello"}],
    "stream": true
  }'
JavaScript (streaming)
const res = await fetch("https://dt.portal.exergynet.org/v1/chat/completions", {
  method: "POST",
  headers: {
    "Authorization": "Bearer sk-exergy-••••••••••••••••",
    "Content-Type": "application/json"
  },
  body: JSON.stringify({
    model: "vanguard-engine",
    messages: [{ role: "user", content: "Hello" }],
    stream: true
  })
});

const reader = res.body.getReader();
// ... read SSE stream
Key Rotation

Rotating your key immediately invalidates the current key and generates a new one. The new key is shown exactly once — store it securely before dismissing.

Environment Variables
EXERGYNET_API_KEY=sk-exergy-••••••••••••••••your API key
EXERGYNET_BASE_URL=https://dt.portal.exergynet.org/v1inference endpoint
OPENAI_BASE_URL=https://dt.portal.exergynet.org/v1if using openai SDK