from zai.client import ZAIClient

# Initialize client with automatic authentication
client = ZAIClient(auto_auth=True)

# Simple chat completion
response = client.simple_chat(
    message="What is the capital of France?",
    model="glm-4.5v"
)
print(response.content)

client = ZAIClient(
    token=None,              # Optional: Bearer token for authentication
    base_url="https://chat.z.ai",  # API base URL
    timeout=180,             # Request timeout in seconds
    auto_auth=True,          # Auto-fetch guest token if no token provided
    verbose=False            # Enable debug output
)

response = client.simple_chat(
    message="Your message here",
    model="glm-4.5v",        # or "0727-360B-API"
    enable_thinking=True,     # Enable thinking mode
    temperature=0.7,          # Control randomness (0.0-2.0)
    top_p=0.9,               # Control diversity (0.0-1.0)
    max_tokens=500           # Maximum response length
)

from zai.models import MCPFeature

chat = client.create_chat(
    title="My Chat",
    models=["glm-4.5v"],
    initial_message="Hello",
    enable_thinking=True,
    features=[
        MCPFeature("mcp", "vibe-coding", "hidden")
    ]
)

# Stream completion with custom chat
for chunk in client.stream_completion(
    chat_id="your-chat-id",
    messages=[
        {"role": "user", "content": "Tell me a story"}
    ],
    model="0727-360B-API",
    enable_thinking=True
):
    if chunk.phase == "answer":
        print(chunk.delta_content, end="")

# Get complete response without streaming
response = client.complete_chat(
    chat_id="your-chat-id",
    messages=[
        {"role": "user", "content": "Explain quantum computing"}
    ],
    model="glm-4.5v"
)
print(response.content)
print(response.thinking)  # If thinking mode was enabled

from zai.custom_models import get_preset, list_presets

# List available presets
print(list_presets())
# Output: ['creative', 'code', 'balanced', 'research', 'brainstorm', 'conservative']

# Get a preset configuration
preset = get_preset('code')
# Returns optimized parameters for code generation

response = client.simple_chat(
    message="Write a creative story",
    model="glm-4.5v",
    temperature=1.2,  # Higher for more creativity
    top_p=0.95,       # Higher for more diversity
    max_tokens=2000   # Longer responses
)

from zai.core.exceptions import ZAIError

try:
    response = client.simple_chat(
        message="Hello",
        model="glm-4.5v"
    )
except ZAIError as e:
    print(f"API Error: {e}")
except Exception as e:
    print(f"Unexpected error: {e}")

# Enable verbose output to see API requests and responses
client = ZAIClient(auto_auth=True, verbose=True)

# Use your own authentication token
client = ZAIClient(
    token="your-bearer-token",
    auto_auth=False  # Disable automatic token fetching
)