Quickstart

Install the OpenAI SDK and point it at the Super AI Engineer LLM endpoint.

pip install openai

from openai import OpenAI

client = OpenAI(
    base_url="https://api.example.com/v1",
    api_key="YOUR_API_KEY"
)

response = client.chat.completions.create(
    model="thai-llm-chat-v0.1",
    messages=[
        {"role": "user", "content": "ช่วยอธิบาย Machine Learning เป็นภาษาไทยง่าย ๆ"}
    ],
    temperature=0.7
)

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

API Reference

Base URL: https://api.example.com/v1

• POST /v1/chat/completions — chat completion (stream optional)
• GET /v1/models — list available models

Authentication

Pass your key as a Bearer token. Keep it server-side; never expose it in a browser.

Authorization: Bearer YOUR_API_KEY

Chat Completion

cURL example:

curl https://api.example.com/v1/chat/completions \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "thai-llm-chat-v0.1",
    "messages": [
      {"role": "user", "content": "สวัสดีครับ"}
    ],
    "temperature": 0.7
  }'

Streaming

Set stream=True to receive tokens as they are generated.

stream = client.chat.completions.create(
    model="thai-llm-chat-v0.1",
    messages=[{"role": "user", "content": "เล่านิทานสั้น ๆ"}],
    stream=True
)

for chunk in stream:
    delta = chunk.choices[0].delta.content or ""
    print(delta, end="", flush=True)

Deployment

Serve with vLLM / TGI on B200 or LANTA, then switch the portal to real mode.

# Serve with vLLM (OpenAI-compatible) on B200 / LANTA
python -m vllm.entrypoints.openai.api_server \
  --model thai-llm/thai-llm-chat-v0.1 \
  --served-model-name thai-llm-chat-v0.1 \
  --port 8000

# Point the portal at it (.env)
# LLM_API_MODE=real
# LLM_API_BASE_URL=http://your-b200-host:8000
# LLM_API_KEY=your-key

Model Limitations

• This is an early version — outputs may contain errors or hallucinations.
• Not intended for medical, legal, or financial advice.
• Knowledge is limited to its training data cutoff.
• Always review generated content before production use.