Completions

Generate sovereign text completions using the Axon language models. The API follows the same message-based format as OpenAI Chat Completions — making it a near-drop-in for existing integrations that need to move to a data-sovereign model.

Create a completion

bash

POST /v1/axon/completions

Name	Type	Required	Description
model	string	Yes	axon-sovereign-1 \| axon-sovereign-1-mini
messages	array	Yes	Array of { role, content } objects. Roles: system \| user \| assistant.
data_residency	string	No	au \| us \| eu \| uk \| sg. Defaults to "au".
max_tokens	integer	No	Max output tokens. Defaults to 2048.
temperature	number	No	0–2. Lower = more deterministic. Default 1.
top_p	number	No	Nucleus sampling. Default 1.
system	string	No	Shorthand system prompt (alternative to a messages entry with role "system").
stream	boolean	No	Server-sent event stream. Default false.
knowledge_base_id	string	No	Attach a knowledge base for automatic RAG context injection.

Basic completion

bash

curl -X POST https://api.hldgroup.org/v1/axon/completions \
  -H "x-internal-secret: <key>" \
  -H "x-tenant-id: ten_01hxyz" \
  -H "x-user-id: usr_01hxyz" \
  -H "x-platform-role: tenant-standard-user" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "axon-sovereign-1",
    "data_residency": "au",
    "messages": [
      {
        "role": "system",
        "content": "You are a concise security analyst assistant for HLD. Answer in plain English."
      },
      {
        "role": "user",
        "content": "Summarise the key steps in a ransomware incident response."
      }
    ],
    "max_tokens": 512,
    "temperature": 0.3
  }'

json

{
  "data": {
    "id": "cmp_01hxyz",
    "object": "chat.completion",
    "model": "axon-sovereign-1",
    "data_residency": "au",
    "sovereign": true,
    "created_at": "2025-06-01T10:00:00Z",
    "choices": [
      {
        "index": 0,
        "message": {
          "role": "assistant",
          "content": "Key ransomware response steps: ..."
        },
        "finish_reason": "stop"
      }
    ],
    "usage": {
      "prompt_tokens": 86,
      "completion_tokens": 201,
      "total_tokens": 287
    }
  }
}

With knowledge base (auto-RAG)

bash

curl -X POST https://api.hldgroup.org/v1/axon/completions \
  -H "x-internal-secret: <key>" \
  -H "x-tenant-id: ten_01hxyz" \
  -H "x-user-id: usr_01hxyz" \
  -H "x-platform-role: tenant-standard-user" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "axon-sovereign-1",
    "data_residency": "au",
    "knowledge_base_id": "kb_01hxyz",
    "messages": [
      { "role": "user", "content": "What does our policy say about third-party vendor access?" }
    ]
  }'

When knowledge_base_id is provided, Axon automatically embeds the user message, retrieves the top matching chunks from your knowledge base, and injects them as context before generating the response.

Note:Completions with sovereign: true in the response confirm the request was processed entirely within HLD-controlled infrastructure. This field will never be false for Axon requests — it's a hard guarantee, not a flag.

Embeddings Knowledge bases