openlit 0.0.1__py3-none-any.whl

openlit/__helpers.py

@@ -0,0 +1,143 @@
+# pylint: disable=bare-except, broad-exception-caught
+"""
+This module has functions to calculate model costs based on tokens and to fetch pricing information.
+"""
+
+import logging
+import requests
+import tiktoken
+from opentelemetry.trace import Status, StatusCode
+
+# Set up logging
+logger = logging.getLogger(__name__)
+
+def openai_tokens(text, model):
+    """
+    Calculate the number of tokens a given text would take up for a specified model.
+    
+    Args:
+        text (str): The input text to be encoded.
+        model (str): The model identifier used for encoding.
+        
+    Returns:
+        int: The number of tokens the text is encoded into.
+    """
+    try:
+        encoding = tiktoken.encoding_for_model(model)
+    except:
+        encoding = tiktoken.get_encoding("cl100k_base")
+
+    num_tokens = len(encoding.encode(text))
+    return num_tokens
+
+def general_tokens(text):
+    """
+    Calculate the number of tokens a given text would take up.
+    
+    Args:
+        text (str): The input text to be encoded.
+        model (str): The model identifier used for encoding.
+        
+    Returns:
+        int: The number of tokens the text is encoded into.
+    """
+
+    encoding = tiktoken.get_encoding("gpt2")
+
+    num_tokens = len(encoding.encode(text))
+    return num_tokens
+
+def get_chat_model_cost(model, pricing_info, prompt_tokens, completion_tokens):
+    """
+    Retrieve the cost of processing for a given model based on prompt and tokens.
+    
+    Args:
+        model (str): The model identifier.
+        pricing_info (dict): A dictionary containing pricing information for various models.
+        prompt_tokens (int): Number of tokens in the prompt.
+        completion_tokens (int): Number of tokens in the completion if applicable.
+        
+    Returns:
+        float: The calculated cost for the operation.
+    """
+    try:
+        cost = ((prompt_tokens / 1000) * pricing_info["chat"][model]["promptPrice"]) + \
+            ((completion_tokens / 1000) * pricing_info["chat"][model]["completionPrice"])
+    except:
+        cost = 0
+    return cost
+
+def get_embed_model_cost(model, pricing_info, prompt_tokens):
+    """
+    Retrieve the cost of processing for a given model based on prompt tokens.
+    
+    Args:
+        model (str): The model identifier.
+        pricing_info (dict): A dictionary containing pricing information for various models.
+        prompt_tokens (int): Number of tokens in the prompt.
+        
+    Returns:
+        float: The calculated cost for the operation.
+    """
+    try:
+        cost = (prompt_tokens / 1000) * pricing_info["embeddings"][model]
+    except:
+        cost = 0
+    return cost
+
+def get_image_model_cost(model, pricing_info, size, quality):
+    """
+    Retrieve the cost of processing for a given model based on image size and quailty.
+    
+    Args:
+        model (str): The model identifier.
+        pricing_info (dict): A dictionary containing pricing information for various models.
+        size (str): Size of the Image.
+        quality (int): Quality of the Image.
+        
+    Returns:
+        float: The calculated cost for the operation.
+    """
+    try:
+        cost = pricing_info["images"][model][quality][size]
+    except:
+        cost = 0
+    return cost
+
+def get_audio_model_cost(model, pricing_info, prompt):
+    """
+    Retrieve the cost of processing for a given model based on prompt.
+    
+    Args:
+        model (str): The model identifier.
+        pricing_info (dict): A dictionary containing pricing information for various models.
+        prompt (str): Prompt to the LLM Model
+        
+    Returns:
+        float: The calculated cost for the operation.
+    """
+    try:
+        cost = (len(prompt) / 1000) * pricing_info["audio"][model]
+    except:
+        cost = 0
+    return cost
+
+def fetch_pricing_info():
+    """Fetches pricing information from a specified URL."""
+    pricing_url = "https://raw.githubusercontent.com/dokulabs/doku/main/assets/pricing.json"
+    try:
+         # Set a timeout of 10 seconds for both the connection and the read
+        response = requests.get(pricing_url, timeout=20)
+        response.raise_for_status()
+        return response.json()
+    except requests.HTTPError as http_err:
+        logger.error("HTTP error occured while fetching pricing info: %s", http_err)
+    except Exception as err:
+        logger.error("Unexpected error occurred while fetching pricing info: %s", err)
+    return {}
+
+def handle_exception(span,e):
+    """Handles Exception when LLM Function fails or trace creation fails."""
+    # Record the exception details within the span
+    span.record_exception(e)
+    span.set_status(Status(StatusCode.ERROR))

openlit/__init__.py

@@ -0,0 +1,184 @@
+"""
+The __init__.py module for the openLIT package.
+This module sets up the openLIT configuration and instrumentation for various
+large language models (LLMs).
+"""
+from typing import Optional, Dict, Any
+import logging
+
+# Import internal modules for setting up tracing and fetching pricing info.
+from openlit.otel.tracing import setup_tracing
+from openlit.otel.metrics import setup_meter
+from openlit.__helpers import fetch_pricing_info
+
+# Instrumentors for various large language models.
+from openlit.instrumentation.openai import OpenAIInstrumentor
+from openlit.instrumentation.anthropic import AnthropicInstrumentor
+from openlit.instrumentation.cohere import CohereInstrumentor
+from openlit.instrumentation.mistral import MistralInstrumentor
+from openlit.instrumentation.langchain import LangChainInstrumentor
+from openlit.instrumentation.chroma import ChromaInstrumentor
+from openlit.instrumentation.pinecone import PineconeInstrumentor
+from openlit.instrumentation.transformers import TransformersInstrumentor
+
+# Set up logging for error and information messages.
+logger = logging.getLogger(__name__)
+
+class OpenlitConfig:
+    """
+    A Singleton Configuration class for openLIT.
+    
+    This class maintains a single instance of configuration settings including
+    environment details, application name, and tracing information throughout the openLIT package.
+    
+    Attributes:
+        environment (str): Deployment environment of the application.
+        application_name (str): Name of the application using openLIT.
+        pricing_info (Dict[str, Any]): Pricing information.
+        tracer (Optional[Any]): Tracer instance for OpenTelemetry.
+        otlp_endpoint (Optional[str]): Endpoint for OTLP.
+        otlp_headers (Optional[Dict[str, str]]): Headers for OTLP.
+        disable_batch (bool): Flag to disable batch span processing in tracing.
+        trace_content (bool): Flag to enable or disable tracing of content.
+    """
+    _instance = None
+
+    def __new__(cls):
+        """Ensures that only one instance of the configuration exists."""
+        if cls._instance is None:
+            cls._instance = super(OpenlitConfig, cls).__new__(cls)
+            cls.reset_to_defaults()
+        return cls._instance
+
+    @classmethod
+    def reset_to_defaults(cls):
+        """Resets configuration to default values."""
+        cls.environment = "default"
+        cls.application_name = "default"
+        cls.pricing_info = fetch_pricing_info()
+        cls.tracer = None
+        cls.metrics_dict = {}
+        cls.otlp_endpoint = None
+        cls.otlp_headers = None
+        cls.disable_batch = False
+        cls.trace_content = True
+        cls.disable_metrics = False
+
+    @classmethod
+    def update_config(cls, environment, application_name, tracer, otlp_endpoint,
+                      otlp_headers, disable_batch, trace_content, metrics_dict, disable_metrics):
+        """
+        Updates the configuration based on provided parameters.
+
+        Args:
+            environment (str): Deployment environment.
+            application_name (str): Application name.
+            tracer: Tracer instance.
+            meter: Metric Instance
+            otlp_endpoint (str): OTLP endpoint.
+            otlp_headers (Dict[str, str]): OTLP headers.
+            disable_batch (bool): Disable batch span processing flag.
+            trace_content (bool): Enable or disable content tracing.
+        """
+        cls.environment = environment
+        cls.application_name = application_name
+        cls.pricing_info = fetch_pricing_info()
+        cls.tracer = tracer
+        cls.metrics_dict = metrics_dict
+        cls.otlp_endpoint = otlp_endpoint
+        cls.otlp_headers = otlp_headers
+        cls.disable_batch = disable_batch
+        cls.trace_content = trace_content
+        cls.disable_metrics = disable_metrics
+
+def init(environment="default", application_name="default", tracer=None, otlp_endpoint=None,
+         otlp_headers=None, disable_batch=False, trace_content=True, disabled_instrumentors=None,
+         meter=None, disable_metrics=False):
+    """
+    Initializes the openLIT configuration and setups tracing.
+    
+    This function sets up the openLIT environment with provided configurations 
+    and initializes instrumentors for tracing.
+    
+    Args:
+        environment (str): Deployment environment.
+        application_name (str): Application name.
+        tracer: Tracer instance (Optional).
+        meter: OpenTelemetry Metrics Instance (Optional).
+        otlp_endpoint (str): OTLP endpoint for exporter (Optional).
+        otlp_headers (Dict[str, str]): OTLP headers for exporter (Optional).
+        disable_batch (bool): Flag to disable batch span processing (Optional).
+        trace_content (bool): Flag to trace content (Optional).
+        disabled_instrumentors (List[str]): Optional. List of instrumentor names to disable.
+        disable_metrics (bool): Flag to disable metrics (Optional)
+    """
+    disabled_instrumentors = disabled_instrumentors if disabled_instrumentors else []
+
+    # Check for invalid instrumentor names
+    valid_instruments = {
+        "openai", "anthropic", "langchain",
+        "cohere", "mistral", "chroma",
+        "pinecone", "transformers"
+    }
+    invalid_instrumentors = set(disabled_instrumentors) - valid_instruments
+    for invalid_name in invalid_instrumentors:
+        logger.warning("Invalid instrumentor name detected and ignored: '%s'", invalid_name)
+
+    try:
+        # Retrieve or create the single configuration instance.
+        config = OpenlitConfig()
+
+        # Setup tracing based on the provided or default configuration.
+        tracer = setup_tracing(
+            application_name=application_name,
+            environment=environment, tracer=tracer,
+            otlp_endpoint=otlp_endpoint, otlp_headers=otlp_headers,
+            disable_batch=disable_batch
+        )
+
+        if not tracer:
+            logger.error("openLIT tracing setup failed. Tracing will not be available.")
+            return
+
+        # Setup meter and receive metrics_dict instead of meter
+        metrics_dict = setup_meter(application_name=application_name,
+                                   environment=environment, meter=meter,
+                                   otlp_endpoint=otlp_endpoint, otlp_headers=otlp_headers)
+
+        if not metrics_dict:
+            logger.error("openLIT metrics setup failed. Metrics will not be available.")
+            return
+
+        # Update global configuration with the provided settings.
+        config.update_config(environment, application_name, tracer, otlp_endpoint,
+                             otlp_headers, disable_batch, trace_content,
+                             metrics_dict, disable_metrics)
+
+        # Map instrumentor names to their instances
+        instrumentor_instances = {
+            "openai": OpenAIInstrumentor(),
+            "anthropic": AnthropicInstrumentor(),
+            "cohere": CohereInstrumentor(),
+            "mistral": MistralInstrumentor(),
+            "langchain": LangChainInstrumentor(),
+            "chroma": ChromaInstrumentor(),
+            "pinecone": PineconeInstrumentor(),
+            "transformers": TransformersInstrumentor()
+        }
+
+        # Initialize and instrument only the enabled instrumentors
+        for name, instrumentor in instrumentor_instances.items():
+            if name not in disabled_instrumentors:
+                instrumentor.instrument(
+                    environment=config.environment,
+                    application_name=config.application_name,
+                    tracer=config.tracer,
+                    pricing_info=config.pricing_info,
+                    trace_content=config.trace_content,
+                    metrics_dict=config.metrics_dict,
+                    disable_metrics=config.disable_metrics
+                )
+
+    # pylint: disable=broad-exception-caught
+    except Exception as e:
+        logger.error("Error during openLIT initialization: %s", e)

openlit/instrumentation/anthropic/__init__.py

@@ -0,0 +1,50 @@
+# pylint: disable=useless-return, bad-staticmethod-argument, disable=duplicate-code
+"""Initializer of Auto Instrumentation of Anthropic Functions"""
+
+from typing import Collection
+import importlib.metadata
+from opentelemetry.instrumentation.instrumentor import BaseInstrumentor
+from wrapt import wrap_function_wrapper
+
+from openlit.instrumentation.anthropic.anthropic import messages
+from openlit.instrumentation.anthropic.async_anthropic import async_messages
+
+_instruments = ("anthropic >= 0.21.0",)
+
+class AnthropicInstrumentor(BaseInstrumentor):
+    """
+    An instrumentor for Anthropic's client library.
+    """
+
+    def instrumentation_dependencies(self) -> Collection[str]:
+        return _instruments
+
+    def _instrument(self, **kwargs):
+        application_name = kwargs.get("application_name", "default_application")
+        environment = kwargs.get("environment", "default_environment")
+        tracer = kwargs.get("tracer")
+        metrics = kwargs.get("metrics_dict")
+        pricing_info = kwargs.get("pricing_info", {})
+        trace_content = kwargs.get("trace_content", False)
+        disable_metrics = kwargs.get("disable_metrics")
+        version = importlib.metadata.version("anthropic")
+
+        #sync
+        wrap_function_wrapper(
+            "anthropic.resources.messages",  
+            "Messages.create",  
+            messages("anthropic.messages", version, environment, application_name,
+                     tracer, pricing_info, trace_content, metrics, disable_metrics),
+        )
+
+        #async
+        wrap_function_wrapper(
+            "anthropic.resources.messages",  
+            "AsyncMessages.create",  
+            async_messages("anthropic.messages", version, environment, application_name,
+                            tracer, pricing_info, trace_content, metrics, disable_metrics),
+        )
+
+    def _uninstrument(self, **kwargs):
+        # Proper uninstrumentation logic to revert patched methods
+        pass

openlit/instrumentation/anthropic/anthropic.py

@@ -0,0 +1,291 @@
+# pylint: disable=duplicate-code, broad-exception-caught, too-many-statements, unused-argument
+"""
+Module for monitoring Anthropic API calls.
+"""
+
+import logging
+from opentelemetry.trace import SpanKind, Status, StatusCode
+from opentelemetry.sdk.resources import TELEMETRY_SDK_NAME
+from openlit.__helpers import get_chat_model_cost, handle_exception
+from openlit.semcov import SemanticConvetion
+
+# Initialize logger for logging potential issues and operations
+logger = logging.getLogger(__name__)
+
+def messages(gen_ai_endpoint, version, environment, application_name, tracer,
+             pricing_info, trace_content, metrics, disable_metrics):
+    """
+    Generates a telemetry wrapper for messages to collect metrics.
+
+    Args:
+        gen_ai_endpoint: Endpoint identifier for logging and tracing.
+        version: Version of the monitoring package.
+        environment: Deployment environment (e.g., production, staging).
+        application_name: Name of the application using the OpenAI API.
+        tracer: OpenTelemetry tracer for creating spans.
+        pricing_info: Information used for calculating the cost of OpenAI usage.
+        trace_content: Flag indicating whether to trace the actual content.
+
+    Returns:
+        A function that wraps the chat method to add telemetry.
+    """
+
+    def wrapper(wrapped, instance, args, kwargs):
+        """
+        Wraps the 'messages' API call to add telemetry.
+
+        This collects metrics such as execution time, cost, and token usage, and handles errors
+        gracefully, adding details to the trace for observability.
+
+        Args:
+            wrapped: The original 'messages' method to be wrapped.
+            instance: The instance of the class where the original method is defined.
+            args: Positional arguments for the 'messages' method.
+            kwargs: Keyword arguments for the 'messages' method.
+
+        Returns:
+            The response from the original 'messages' method.
+        """
+
+        # Check if streaming is enabled for the API call
+        streaming = kwargs.get("stream", False)
+
+        # pylint: disable=no-else-return
+        if streaming:
+            # Special handling for streaming response to accommodate the nature of data flow
+            def stream_generator():
+                with tracer.start_as_current_span(gen_ai_endpoint, kind= SpanKind.CLIENT) as span:
+                    # Placeholder for aggregating streaming response
+                    llmresponse = ""
+
+                    # Loop through streaming events capturing relevant details
+                    for event in wrapped(*args, **kwargs):
+
+                        # Collect message IDs and input token from events
+                        if event.type == "message_start":
+                            response_id = event.message.id
+                            prompt_tokens = event.message.usage.input_tokens
+
+                        # Aggregate response content
+                        if event.type == "content_block_delta":
+                            llmresponse += event.delta.text
+
+                        # Collect output tokens and stop reason from events
+                        if event.type == "message_delta":
+                            completion_tokens = event.usage.output_tokens
+                            finish_reason = event.delta.stop_reason
+                        yield event
+
+                    # Handling exception ensure observability without disrupting operation
+                    try:
+                        # Format 'messages' into a single string
+                        message_prompt = kwargs.get("messages", "")
+                        formatted_messages = []
+                        for message in message_prompt:
+                            role = message["role"]
+                            content = message["content"]
+
+                            if isinstance(content, list):
+                                content_str = ", ".join(
+                                    # pylint: disable=line-too-long
+                                    f'{item["type"]}: {item["text"] if "text" in item else item["image_url"]}'
+                                    if "type" in item else f'text: {item["text"]}'
+                                    for item in content
+                                )
+                                formatted_messages.append(f"{role}: {content_str}")
+                            else:
+                                formatted_messages.append(f"{role}: {content}")
+                        prompt = "\n".join(formatted_messages)
+
+                        # Calculate cost of the operation
+                        cost = get_chat_model_cost(
+                            kwargs.get("model", "claude-3-sonnet-20240229"),
+                            pricing_info, prompt_tokens, completion_tokens
+                        )
+
+                        # Set Span attributes
+                        span.set_attribute(TELEMETRY_SDK_NAME, "openlit")
+                        span.set_attribute(SemanticConvetion.GEN_AI_SYSTEM,
+                                            SemanticConvetion.GEN_AI_SYSTEM_ANTHROPIC)
+                        span.set_attribute(SemanticConvetion.GEN_AI_TYPE,
+                                            SemanticConvetion.GEN_AI_TYPE_CHAT)
+                        span.set_attribute(SemanticConvetion.GEN_AI_ENDPOINT,
+                                            gen_ai_endpoint)
+                        span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_ID,
+                                            response_id)
+                        span.set_attribute(SemanticConvetion.GEN_AI_ENVIRONMENT,
+                                            environment)
+                        span.set_attribute(SemanticConvetion.GEN_AI_APPLICATION_NAME,
+                                            application_name)
+                        span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MODEL,
+                                            kwargs.get("model", "claude-3-sonnet-20240229"))
+                        span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MAX_TOKENS,
+                                            kwargs.get("max_tokens", ""))
+                        span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_IS_STREAM,
+                                            True)
+                        span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TEMPERATURE,
+                                            kwargs.get("temperature", 1.0))
+                        span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TOP_P,
+                                            kwargs.get("top_p", ""))
+                        span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TOP_K,
+                                            kwargs.get("top_k", ""))
+                        span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_FINISH_REASON,
+                                            finish_reason)
+                        span.set_attribute(SemanticConvetion.GEN_AI_USAGE_PROMPT_TOKENS,
+                                            prompt_tokens)
+                        span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COMPLETION_TOKENS,
+                                            completion_tokens)
+                        span.set_attribute(SemanticConvetion.GEN_AI_USAGE_TOTAL_TOKENS,
+                                            prompt_tokens + completion_tokens)
+                        span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
+                                            cost)
+                        if trace_content:
+                            span.set_attribute(SemanticConvetion.GEN_AI_CONTENT_PROMPT,
+                                                prompt)
+                            span.set_attribute(SemanticConvetion.GEN_AI_CONTENT_COMPLETION,
+                                                llmresponse)
+
+                        span.set_status(Status(StatusCode.OK))
+
+                        if disable_metrics is False:
+                            attributes = {
+                                TELEMETRY_SDK_NAME:
+                                    "openlit",
+                                SemanticConvetion.GEN_AI_APPLICATION_NAME:
+                                    application_name,
+                                SemanticConvetion.GEN_AI_SYSTEM:
+                                    SemanticConvetion.GEN_AI_SYSTEM_ANTHROPIC,
+                                SemanticConvetion.GEN_AI_ENVIRONMENT:
+                                    environment,
+                                SemanticConvetion.GEN_AI_TYPE:
+                                    SemanticConvetion.GEN_AI_TYPE_CHAT,
+                                SemanticConvetion.GEN_AI_REQUEST_MODEL:
+                                    kwargs.get("model", "claude-3-sonnet-20240229")
+                            }
+
+                            metrics["genai_requests"].add(1, attributes)
+                            metrics["genai_total_tokens"].add(
+                                prompt_tokens + completion_tokens, attributes
+                            )
+                            metrics["genai_completion_tokens"].add(completion_tokens, attributes)
+                            metrics["genai_prompt_tokens"].add(prompt_tokens, attributes)
+                            metrics["genai_cost"].record(cost, attributes)
+
+                    except Exception as e:
+                        handle_exception(span, e)
+                        logger.error("Error in trace creation: %s", e)
+
+            return stream_generator()
+
+        # Handling for non-streaming responses
+        else:
+            with tracer.start_as_current_span(gen_ai_endpoint, kind=SpanKind.CLIENT) as span:
+                response = wrapped(*args, **kwargs)
+
+                try:
+                    # Format 'messages' into a single string
+                    message_prompt = kwargs.get("messages", "")
+                    formatted_messages = []
+                    for message in message_prompt:
+                        role = message["role"]
+                        content = message["content"]
+
+                        if isinstance(content, list):
+                            content_str = ", ".join(
+                                # pylint: disable=line-too-long
+                                f'{item["type"]}: {item["text"] if "text" in item else item["image_url"]}'
+                                if "type" in item else f'text: {item["text"]}'
+                                for item in content
+                            )
+                            formatted_messages.append(f"{role}: {content_str}")
+                        else:
+                            formatted_messages.append(f"{role}: {content}")
+                    prompt = "\n".join(formatted_messages)
+
+                    # Calculate cost of the operation
+                    cost = get_chat_model_cost(kwargs.get("model", "claude-3-sonnet-20240229"),
+                                                pricing_info, response.usage.input_tokens,
+                                                response.usage.output_tokens)
+
+                    # Set Span attribues
+                    span.set_attribute(TELEMETRY_SDK_NAME, "openlit")
+                    span.set_attribute(SemanticConvetion.GEN_AI_SYSTEM,
+                                        SemanticConvetion.GEN_AI_SYSTEM_ANTHROPIC)
+                    span.set_attribute(SemanticConvetion.GEN_AI_TYPE,
+                                        SemanticConvetion.GEN_AI_TYPE_CHAT)
+                    span.set_attribute(SemanticConvetion.GEN_AI_ENDPOINT,
+                                        gen_ai_endpoint)
+                    span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_ID,
+                                        response.id)
+                    span.set_attribute(SemanticConvetion.GEN_AI_ENVIRONMENT,
+                                        environment)
+                    span.set_attribute(SemanticConvetion.GEN_AI_APPLICATION_NAME,
+                                        application_name)
+                    span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MODEL,
+                                        kwargs.get("model", "claude-3-sonnet-20240229"))
+                    span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MAX_TOKENS,
+                                        kwargs.get("max_tokens", ""))
+                    span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_IS_STREAM,
+                                        False)
+                    span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TEMPERATURE,
+                                        kwargs.get("temperature", 1.0))
+                    span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TOP_P,
+                                        kwargs.get("top_p", ""))
+                    span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TOP_K,
+                                        kwargs.get("top_k", ""))
+                    span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_FINISH_REASON,
+                                        response.stop_reason)
+                    span.set_attribute(SemanticConvetion.GEN_AI_USAGE_PROMPT_TOKENS,
+                                        response.usage.input_tokens)
+                    span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COMPLETION_TOKENS,
+                                        response.usage.output_tokens)
+                    span.set_attribute(SemanticConvetion.GEN_AI_USAGE_TOTAL_TOKENS,
+                                        response.usage.input_tokens +
+                                        response.usage.output_tokens)
+                    span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
+                                        cost)
+                    if trace_content:
+                        span.set_attribute(SemanticConvetion.GEN_AI_CONTENT_PROMPT,
+                                            prompt)
+                        # pylint: disable=line-too-long
+                        span.set_attribute(SemanticConvetion.GEN_AI_CONTENT_COMPLETION, response.content[0].text if response.content else "")
+
+                    span.set_status(Status(StatusCode.OK))
+
+                    if disable_metrics is False:
+                        attributes = {
+                            TELEMETRY_SDK_NAME:
+                                "openlit",
+                            SemanticConvetion.GEN_AI_APPLICATION_NAME:
+                                application_name,
+                            SemanticConvetion.GEN_AI_SYSTEM:
+                                SemanticConvetion.GEN_AI_SYSTEM_ANTHROPIC,
+                            SemanticConvetion.GEN_AI_ENVIRONMENT:
+                                environment,
+                            SemanticConvetion.GEN_AI_TYPE:
+                                SemanticConvetion.GEN_AI_TYPE_CHAT,
+                            SemanticConvetion.GEN_AI_REQUEST_MODEL:
+                                kwargs.get("model", "claude-3-sonnet-20240229")
+                        }
+
+                        metrics["genai_requests"].add(1, attributes)
+                        metrics["genai_total_tokens"].add(
+                            response.usage.input_tokens +
+                            response.usage.output_tokens, attributes)
+                        metrics["genai_completion_tokens"].add(
+                            response.usage.output_tokens, attributes)
+                        metrics["genai_prompt_tokens"].add(
+                            response.usage.input_tokens, attributes)
+                        metrics["genai_cost"].record(cost, attributes)
+
+                    # Return original response
+                    return response
+
+                except Exception as e:
+                    handle_exception(span, e)
+                    logger.error("Error in trace creation: %s", e)
+
+                    # Return original response
+                    return response
+
+    return wrapper