GPT ILM Chat Completions

from openai import OpenAI

client = OpenAI(
    api_key="sk-XXX",
    base_url="https://mavi-backend.memories.ai/serve/api/v2/ilm"
)

def call_my_ilm():
    resp = client.chat.completions.create(
        model="gpt:gpt-4o",
        messages=[
            {"role": "system", "content": "You are a multimodal assistant. Only output JSON, do not output any extra text."},
            {
                "role": "user",
                "content": [
                    {
                        "type": "text",
                        "text": """
Please analyze the image and strictly output the following JSON structure (all fields must be present, use null or empty array for missing values):

{
  "title": string,
  "summary": string,
  "objects": [
    {"name": string, "count": integer, "confidence": number}
  ],
  "scene": string,
  "warnings": [string]
}

Note:
- Only output JSON
- No Markdown
- No explanation of the process
"""
                    },
                    {
                        "type": "image_url",
                        "image_url": {
                            "url": "https://storage.googleapis.com/memories-test-data/gun5.png"  # base64 or url
                        }
                    }
                ]
            }
        ],
        response_format={"type": "json_object"},  # Force JSON output format
        temperature=0,  # Controls randomness: 0.0-2.0, 0 = deterministic
        max_tokens=1000,  # Maximum number of tokens to generate
        top_p=1.0,  # Nucleus sampling: 0.0-1.0, consider tokens with top_p probability mass
        frequency_penalty=0.0,  # -2.0 to 2.0, reduces repetition of frequent tokens
        presence_penalty=0.0,  # -2.0 to 2.0, increases likelihood of new topics
        n=1,  # Number of completions to generate
        stream=False,  # Whether to stream the response
        stop=None  # Stop sequences (list of strings)
    )
    return resp

# Usage example
result = call_my_ilm()
print(result)