Helm Completions API

OpenAI-compatible completions endpoint backed by MetalMind PhotonicGPT-v2 (14M params). Sovereign inference — no external APIs.

Endpoints

• GET /v1/models — list available models
• POST /v1/chat/completions — chat completions
• POST /v1/infer — raw inference

Chat Completions

POST /v1/chat/completions
Content-Type: application/json

{
  "model": "photonic-gpt-v2",
  "messages": [
    {"role": "user", "content": "What can you build?"}
  ],
  "max_tokens": 200,
  "temperature": 0.85
}

OpenAI SDK Usage

from openai import OpenAI

client = OpenAI(
    base_url="https://completions.helmcorp.cc/v1",
    api_key="sovereign"  # any value
)

response = client.chat.completions.create(
    model="photonic-gpt-v2",
    messages=[{"role": "user", "content": "Hello"}]
)
print(response.choices[0].message.content)

Model

• Engine: MetalMind PhotonicGPT-v2
• Parameters: 14M
• Inference: sovereign local — no cloud, no external APIs
• Stack: flowistan_tls → weyland :7720 → metalmind_cpu