LiteLLM API - Documentation

The LiteLLM API allows you to interact with large language models (LLM) from your tools or IDE. It simplifies integration with various providers (OpenAI, Ollama, etc.).

Base URL: https://api.ia.limos.fr

API Key Management

You can create an API key in your ia.limos.fr space → "My keys" tab.

For interns, generate a temporary key valid for 30 days (renewable). This key is linked to your account and can be deleted at any time.

---

Basic Configuration

For services using an LLM (e.g., chatbots, AI assistants), specify: - URL: https://api.ia.limos.fr/v1 - API Key: the one generated on ia.limos.fr

---

Basic Commands

List available models

Bash

curl -X GET "https://api.ia.limos.fr/v1/models" -H "Authorization: Bearer API_KEY"

Python

import requests

headers = {
    "Authorization": "Bearer API_KEY"  # Replace with your key
}

response = requests.get("https://api.ia.limos.fr/v1/models", headers=headers)
print(response.json())

---

Query for chat completion

Bash

curl -X POST "https://api.ia.limos.fr/v1/chat/completions" \
  -H "Authorization: Bearer API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
      "model": "general_nothink",
      "messages": [
          {"role": "system", "content": "You are a clear and concise assistant."},
          {"role": "user", "content": "Tell me the capital of France."}
      ],
      "max_tokens": 50,
      "temperature": 0.7
    }'

Python

import requests
import json

data = {
    "model": "general_nothink",
    "messages": [
        {"role": "system", "content": "You are a clear and concise assistant."},
        {"role": "user", "content": "Tell me the capital of France."}
    ],
    "max_tokens": 50,
    "temperature": 0.7
}

headers = {
    "Authorization": "Bearer API_KEY",  # Replace with your key
    "Content-Type": "application/json"
}

response = requests.post(
    "https://api.ia.limos.fr/v1/chat/completions",
    headers=headers,
    data=json.dumps(data)
)
print(response.json())

Query for an embedding (vector representation of text)

Bash

curl -X POST "https://api.ia.limos.fr/v1/embeddings" \
  -H "Authorization: Bearer API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
      "model": "embedding",
      "input": "Hello, how are you?"
  }'

Python

import requests
import json

data = {
    "model": "embedding-model",
    "input": "Hello, how are you?"
}

headers = {
    "Authorization": "Bearer API_KEY",  # Replace with your key
    "Content-Type": "application/json"
}

response = requests.post(
    "https://api.ia.limos.fr/v1/embeddings",
    headers=headers,
    data=json.dumps(data)
)
print(response.json())

---

Additional Notes

• Required headers:
• Authorization: Bearer <your_key>
• Content-Type: application/json (for POST requests).
• Adjustable parameters:
• temperature (0.0 to 1.0): controls the randomness of responses.
• max_tokens: limits the length of the response.