Skip to main content
Source Code: src/gaia/llm/
Import: from gaia.llm.llm_client import LLMClient
Detailed Spec: spec/llm-client Purpose: Unified client for local and cloud LLM providers.

​
LLM Client

from gaia.llm.llm_client import LLMClient

# Local LLM (Lemonade server)
llm = LLMClient(
    base_url="http://localhost:8000/api/v1",
    system_prompt="You are a helpful assistant"
)

# Generate response
response = llm.generate(
    prompt="What is AI?",
    model="Qwen2.5-0.5B-Instruct-CPU",
    max_tokens=200,
    temperature=0.7
)
print(response)

# Streaming response
for chunk in llm.generate(prompt="Tell me a story", stream=True):
    print(chunk, end="", flush=True)

# Chat completions format
messages = [
    {"role": "user", "content": "Hello!"},
    {"role": "assistant", "content": "Hi! How can I help?"},
    {"role": "user", "content": "Tell me about Python"}
]

response = llm.chat_completions(
    messages=messages,
    model="Qwen2.5-0.5B-Instruct-CPU"
)
print(response)

# List available models
models = llm.get_available_models()
for model in models:
    print(f"- {model}")

​
Cloud Providers

from gaia.llm.llm_client import LLMClient

# Claude API
llm_claude = LLMClient(
    use_claude=True,
    claude_model="claude-sonnet-4-20250514",
    system_prompt="You are an expert coder"
)

# OpenAI API
llm_openai = LLMClient(
    use_openai=True,
    system_prompt="You are a helpful assistant"
)

# Use same interface
response = llm_claude.generate("Explain Python decorators")

​
Lemonade Client (AMD-Optimized)

Import: from gaia.llm.lemonade_client import DEFAULT_MODEL_NAME, DEFAULT_LEMONADE_URL 
from gaia.llm.lemonade_client import DEFAULT_MODEL_NAME, DEFAULT_LEMONADE_URL

# Default configuration
MODEL = DEFAULT_MODEL_NAME  # "Qwen2.5-0.5B-Instruct-CPU"
URL = DEFAULT_LEMONADE_URL   # "http://localhost:8000/api/v1"

# Used internally by LLMClient when no provider specified