# PrimeTechnology LLM API — Agent Reference

Base URL: https://ollama.primetechnology.vn
Auth: Authorization: Bearer <your-team-api-key>
Note: Get your API key from /home/deploypython/ollama-api/.env on server05 (TEAM_API_KEYS)

## Available Models
- minimax-m3       (MiniMax M3 cloud)
- kimi-k2.6        (Moonshot Kimi K2.6 cloud)
- glm-5.1          (ZhipuAI GLM-5.1 cloud)
- qwen3.5:397b     (Alibaba Qwen3.5 397B cloud)

## Endpoints

### GET /health
No auth required. Returns service status and model list.

### GET /v1/models
Auth required. Returns OpenAI-format model list.

### POST /v1/chat/completions
Auth required. OpenAI-compatible chat endpoint.
Request body (JSON):
  model       string   required  one of the allowed models above
  messages    array    required  [{role: "user"|"assistant"|"system", content: string}, ...]
  stream      boolean  optional  default false
  temperature float    optional
  top_p       float    optional

Response: OpenAI ChatCompletion object

Example:
  curl -X POST https://ollama.primetechnology.vn/v1/chat/completions \
    -H "Authorization: Bearer pd_team_2026_xK9mQzRvWpLnTjYe" \
    -H "Content-Type: application/json" \
    -d '{"model":"minimax-m3","messages":[{"role":"user","content":"Hello"}]}'

### GET /api/tags
Auth required. Returns Ollama-format model list.

### POST /api/chat
Auth required. Ollama-native chat endpoint.
Request body (JSON):
  model    string   required
  messages array    required  [{role, content}, ...]
  stream   boolean  optional  default false

Response: Ollama chat response object

## SDK Usage

### OpenAI Python SDK
  from openai import OpenAI
  client = OpenAI(
    base_url="https://ollama.primetechnology.vn/v1",
    api_key="pd_team_2026_xK9mQzRvWpLnTjYe"
  )
  resp = client.chat.completions.create(
    model="minimax-m3",
    messages=[{"role": "user", "content": "Hello"}]
  )
  print(resp.choices[0].message.content)

### Ollama Python SDK
  import ollama
  client = ollama.Client(
    host="https://ollama.primetechnology.vn",
    headers={"Authorization": "Bearer pd_team_2026_xK9mQzRvWpLnTjYe"}
  )
  resp = client.chat(model="minimax-m3", messages=[{"role":"user","content":"Hello"}])
  print(resp.message.content)

## Error Codes
  400  Model not in allowlist
  401  Missing or invalid API key
  500  Upstream Ollama cloud error