openlit 1.1.3__py3-none-any.whl → 1.5.0__py3-none-any.whl

openlit/instrumentation/vertexai/async_vertexai.py

@@ -0,0 +1,1047 @@
+# pylint: disable=duplicate-code, broad-exception-caught, too-many-statements, unused-argument, protected-access,  too-many-lines
+"""
+Module for monitoring VertexAI API calls.
+"""
+
+import logging
+import math
+from opentelemetry.trace import SpanKind, Status, StatusCode
+from opentelemetry.sdk.resources import TELEMETRY_SDK_NAME
+from openlit.__helpers import get_chat_model_cost, get_embed_model_cost, handle_exception
+from openlit.semcov import SemanticConvetion
+
+# Initialize logger for logging potential issues and operations
+logger = logging.getLogger(__name__)
+
+def generate_content_async(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.
+    """
+
+    async def wrapper(wrapped, instance, args, kwargs):
+        """
+        Wraps the 'generate_content' 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 'generate_content' method to be wrapped.
+            instance: The instance of the class where the original method is defined.
+            args: Positional arguments for the 'generate_content' method.
+            kwargs: Keyword arguments for the 'generate_content' method.
+
+        Returns:
+            The response from the original 'generate_content' 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
+            async 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
+                    async for event in await wrapped(*args, **kwargs):
+                        llmresponse += str(event.text)
+                        prompt_tokens = event.usage_metadata.prompt_token_count
+                        completion_tokens = event.usage_metadata.candidates_token_count
+                        total_tokens = event.usage_metadata.total_token_count
+                        yield event
+
+                    # Handling exception ensure observability without disrupting operation
+                    try:
+                        prompt = str(args[0][0])
+
+                        model = "/".join(instance._model_name.split("/")[3:])
+
+                        # Calculate cost of the operation
+                        cost = get_chat_model_cost(model,
+                                                    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_VERTEXAI)
+                        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_ENVIRONMENT,
+                                            environment)
+                        span.set_attribute(SemanticConvetion.GEN_AI_APPLICATION_NAME,
+                                            application_name)
+                        span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MODEL,
+                                            model)
+                        span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_IS_STREAM,
+                                            True)
+                        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,
+                                            total_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_VERTEXAI,
+                                SemanticConvetion.GEN_AI_ENVIRONMENT:
+                                    environment,
+                                SemanticConvetion.GEN_AI_TYPE:
+                                    SemanticConvetion.GEN_AI_TYPE_CHAT,
+                                SemanticConvetion.GEN_AI_REQUEST_MODEL:
+                                    model
+                            }
+
+                            metrics["genai_requests"].add(1, attributes)
+                            metrics["genai_total_tokens"].add(
+                                total_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 = await wrapped(*args, **kwargs)
+
+                try:
+                    # Format 'messages' into a single string
+                    prompt = str(args[0][0])
+
+                    model = "/".join(instance._model_name.split("/")[3:])
+                    # Calculate cost of the operation
+                    cost = get_chat_model_cost(model,
+                                                pricing_info,
+                                                response.usage_metadata.prompt_token_count,
+                                                response.usage_metadata.candidates_token_count)
+
+                    # Set Span attribues
+                    span.set_attribute(TELEMETRY_SDK_NAME, "openlit")
+                    span.set_attribute(SemanticConvetion.GEN_AI_SYSTEM,
+                                        SemanticConvetion.GEN_AI_SYSTEM_VERTEXAI)
+                    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_ENVIRONMENT,
+                                        environment)
+                    span.set_attribute(SemanticConvetion.GEN_AI_APPLICATION_NAME,
+                                        application_name)
+                    span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MODEL,
+                                        model)
+                    span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_IS_STREAM,
+                                        False)
+                    span.set_attribute(SemanticConvetion.GEN_AI_USAGE_PROMPT_TOKENS,
+                                        response.usage_metadata.prompt_token_count)
+                    span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COMPLETION_TOKENS,
+                                        response.usage_metadata.candidates_token_count)
+                    span.set_attribute(SemanticConvetion.GEN_AI_USAGE_TOTAL_TOKENS,
+                                        response.usage_metadata.total_token_count)
+                    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.candidates[0].content.parts[0].text)
+
+                    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_VERTEXAI,
+                            SemanticConvetion.GEN_AI_ENVIRONMENT:
+                                environment,
+                            SemanticConvetion.GEN_AI_TYPE:
+                                SemanticConvetion.GEN_AI_TYPE_CHAT,
+                            SemanticConvetion.GEN_AI_REQUEST_MODEL:
+                                model
+                        }
+
+                        metrics["genai_requests"].add(1, attributes)
+                        metrics["genai_total_tokens"].add(
+                            response.usage_metadata.total_token_count, attributes)
+                        metrics["genai_completion_tokens"].add(
+                            response.usage_metadata.candidates_token_count, attributes)
+                        metrics["genai_prompt_tokens"].add(
+                            response.usage_metadata.prompt_token_count, 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
+
+
+def send_message_async(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.
+    """
+
+    async def wrapper(wrapped, instance, args, kwargs):
+        """
+        Wraps the 'generate_content' 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 'generate_content' method to be wrapped.
+            instance: The instance of the class where the original method is defined.
+            args: Positional arguments for the 'generate_content' method.
+            kwargs: Keyword arguments for the 'generate_content' method.
+
+        Returns:
+            The response from the original 'generate_content' 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
+            async 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
+                    async for event in await wrapped(*args, **kwargs):
+                        llmresponse += str(event.text)
+                        prompt_tokens = event.usage_metadata.prompt_token_count
+                        completion_tokens = event.usage_metadata.candidates_token_count
+                        total_tokens = event.usage_metadata.total_token_count
+                        yield event
+
+                    # Handling exception ensure observability without disrupting operation
+                    try:
+                        prompt = args[0][0]
+
+                        model = "/".join(instance._model._model_name.split("/")[3:])
+
+                        # Calculate cost of the operation
+                        cost = get_chat_model_cost(model,
+                                                    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_VERTEXAI)
+                        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_ENVIRONMENT,
+                                            environment)
+                        span.set_attribute(SemanticConvetion.GEN_AI_APPLICATION_NAME,
+                                            application_name)
+                        span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MODEL,
+                                            model)
+                        span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_IS_STREAM,
+                                            True)
+                        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,
+                                            total_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_VERTEXAI,
+                                SemanticConvetion.GEN_AI_ENVIRONMENT:
+                                    environment,
+                                SemanticConvetion.GEN_AI_TYPE:
+                                    SemanticConvetion.GEN_AI_TYPE_CHAT,
+                                SemanticConvetion.GEN_AI_REQUEST_MODEL:
+                                    model
+                            }
+
+                            metrics["genai_requests"].add(1, attributes)
+                            metrics["genai_total_tokens"].add(
+                                total_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 = await wrapped(*args, **kwargs)
+
+                try:
+                    # Format 'messages' into a single string
+                    prompt = args[0]
+
+                    model = "/".join(instance._model._model_name.split("/")[3:])
+                    # Calculate cost of the operation
+                    cost = get_chat_model_cost(model,
+                                                pricing_info,
+                                                response.usage_metadata.prompt_token_count,
+                                                response.usage_metadata.candidates_token_count)
+
+                    # Set Span attribues
+                    span.set_attribute(TELEMETRY_SDK_NAME, "openlit")
+                    span.set_attribute(SemanticConvetion.GEN_AI_SYSTEM,
+                                        SemanticConvetion.GEN_AI_SYSTEM_VERTEXAI)
+                    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_ENVIRONMENT,
+                                        environment)
+                    span.set_attribute(SemanticConvetion.GEN_AI_APPLICATION_NAME,
+                                        application_name)
+                    span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MODEL,
+                                        model)
+                    span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_IS_STREAM,
+                                        False)
+                    span.set_attribute(SemanticConvetion.GEN_AI_USAGE_PROMPT_TOKENS,
+                                        response.usage_metadata.prompt_token_count)
+                    span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COMPLETION_TOKENS,
+                                        response.usage_metadata.candidates_token_count)
+                    span.set_attribute(SemanticConvetion.GEN_AI_USAGE_TOTAL_TOKENS,
+                                        response.usage_metadata.total_token_count)
+                    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.candidates[0].content.parts[0].text)
+
+                    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_VERTEXAI,
+                            SemanticConvetion.GEN_AI_ENVIRONMENT:
+                                environment,
+                            SemanticConvetion.GEN_AI_TYPE:
+                                SemanticConvetion.GEN_AI_TYPE_CHAT,
+                            SemanticConvetion.GEN_AI_REQUEST_MODEL:
+                                model
+                        }
+
+                        metrics["genai_requests"].add(1, attributes)
+                        metrics["genai_total_tokens"].add(
+                            response.usage_metadata.total_token_count, attributes)
+                        metrics["genai_completion_tokens"].add(
+                            response.usage_metadata.candidates_token_count, attributes)
+                        metrics["genai_prompt_tokens"].add(
+                            response.usage_metadata.prompt_token_count, 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
+
+def predict_async(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.
+    """
+
+    async def wrapper(wrapped, instance, args, kwargs):
+        """
+        Wraps the 'generate_content' 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 'generate_content' method to be wrapped.
+            instance: The instance of the class where the original method is defined.
+            args: Positional arguments for the 'generate_content' method.
+            kwargs: Keyword arguments for the 'generate_content' method.
+
+        Returns:
+            The response from the original 'generate_content' method.
+        """
+
+        with tracer.start_as_current_span(gen_ai_endpoint, kind=SpanKind.CLIENT) as span:
+            response = await wrapped(*args, **kwargs)
+
+            try:
+                prompt = args[0]
+
+                model = instance._model_id
+                #pylint: disable=line-too-long
+                prompt_tokens = response._prediction_response.metadata["tokenMetadata"]["inputTokenCount"]["totalTokens"]
+                completion_tokens = response._prediction_response.metadata["tokenMetadata"]["outputTokenCount"]["totalTokens"]
+                total_tokens = prompt_tokens + completion_tokens
+
+                #Calculate cost of the operation
+                cost = get_chat_model_cost(model,
+                                            pricing_info, prompt_tokens,
+                                            completion_tokens)
+
+                # Set Span attribues
+                span.set_attribute(TELEMETRY_SDK_NAME, "openlit")
+                span.set_attribute(SemanticConvetion.GEN_AI_SYSTEM,
+                                    SemanticConvetion.GEN_AI_SYSTEM_VERTEXAI)
+                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_ENVIRONMENT,
+                                    environment)
+                span.set_attribute(SemanticConvetion.GEN_AI_APPLICATION_NAME,
+                                    application_name)
+                span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MODEL,
+                                    model)
+                span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_IS_STREAM,
+                                    False)
+                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,
+                                    total_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.text)
+
+                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_VERTEXAI,
+                        SemanticConvetion.GEN_AI_ENVIRONMENT:
+                            environment,
+                        SemanticConvetion.GEN_AI_TYPE:
+                            SemanticConvetion.GEN_AI_TYPE_CHAT,
+                        SemanticConvetion.GEN_AI_REQUEST_MODEL:
+                            model
+                    }
+
+                    metrics["genai_requests"].add(1, attributes)
+                    metrics["genai_total_tokens"].add(
+                        total_tokens, attributes)
+                    metrics["genai_completion_tokens"].add(
+                        completion_tokens, attributes)
+                    metrics["genai_prompt_tokens"].add(
+                        prompt_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
+
+def predict_streaming_async(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.
+    """
+
+    async def wrapper(wrapped, instance, args, kwargs):
+        """
+        Wraps the 'predict' 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 'predict' method to be wrapped.
+            instance: The instance of the class where the original method is defined.
+            args: Positional arguments for the 'predict' method.
+            kwargs: Keyword arguments for the 'predict' method.
+
+        Returns:
+            The response from the original 'predict' method.
+        """
+
+        # Special handling for streaming response to accommodate the nature of data flow
+        async 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
+                async for event in wrapped(*args, **kwargs):
+                    llmresponse += str(event)
+                    yield event
+
+                # Handling exception ensure observability without disrupting operation
+                try:
+                    prompt = args[0]
+                    llmresponse = llmresponse.split('TextGenerationResponse',
+                                                    maxsplit=1)[0].rstrip()
+
+                    prompt_tokens = math.ceil(len(prompt) / 4)
+                    completion_tokens = math.ceil(len(llmresponse) / 4)
+                    total_tokens = prompt_tokens + completion_tokens
+
+                    model = instance._model_id
+
+                    # Calculate cost of the operation
+                    cost = get_chat_model_cost(model,
+                                                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_VERTEXAI)
+                    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_ENVIRONMENT,
+                                        environment)
+                    span.set_attribute(SemanticConvetion.GEN_AI_APPLICATION_NAME,
+                                        application_name)
+                    span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MODEL,
+                                        model)
+                    span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_IS_STREAM,
+                                        True)
+                    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,
+                                        total_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_VERTEXAI,
+                            SemanticConvetion.GEN_AI_ENVIRONMENT:
+                                environment,
+                            SemanticConvetion.GEN_AI_TYPE:
+                                SemanticConvetion.GEN_AI_TYPE_CHAT,
+                            SemanticConvetion.GEN_AI_REQUEST_MODEL:
+                                model
+                        }
+
+                        metrics["genai_requests"].add(1, attributes)
+                        metrics["genai_total_tokens"].add(
+                            total_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()
+
+    return wrapper
+
+def start_chat_async(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.
+    """
+
+    async def wrapper(wrapped, instance, args, kwargs):
+        """
+        Wraps the 'generate_content' 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 'generate_content' method to be wrapped.
+            instance: The instance of the class where the original method is defined.
+            args: Positional arguments for the 'generate_content' method.
+            kwargs: Keyword arguments for the 'generate_content' method.
+
+        Returns:
+            The response from the original 'generate_content' method.
+        """
+
+        with tracer.start_as_current_span(gen_ai_endpoint, kind=SpanKind.CLIENT) as span:
+            response =  await wrapped(*args, **kwargs)
+
+            try:
+                prompt = args[0]
+
+                model = instance._model._model_id
+
+                #pylint: disable=line-too-long
+                prompt_tokens = response._prediction_response.metadata["tokenMetadata"]["inputTokenCount"]["totalTokens"]
+                completion_tokens = response._prediction_response.metadata["tokenMetadata"]["outputTokenCount"]["totalTokens"]
+                total_tokens = prompt_tokens + completion_tokens
+
+                #Calculate cost of the operation
+                cost = get_chat_model_cost(model,
+                                            pricing_info, prompt_tokens,
+                                            completion_tokens)
+
+                # Set Span attribues
+                span.set_attribute(TELEMETRY_SDK_NAME, "openlit")
+                span.set_attribute(SemanticConvetion.GEN_AI_SYSTEM,
+                                    SemanticConvetion.GEN_AI_SYSTEM_VERTEXAI)
+                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_ENVIRONMENT,
+                                    environment)
+                span.set_attribute(SemanticConvetion.GEN_AI_APPLICATION_NAME,
+                                    application_name)
+                span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MODEL,
+                                    model)
+                span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_IS_STREAM,
+                                    False)
+                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,
+                                    total_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.text)
+
+                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_VERTEXAI,
+                        SemanticConvetion.GEN_AI_ENVIRONMENT:
+                            environment,
+                        SemanticConvetion.GEN_AI_TYPE:
+                            SemanticConvetion.GEN_AI_TYPE_CHAT,
+                        SemanticConvetion.GEN_AI_REQUEST_MODEL:
+                            model
+                    }
+
+                    metrics["genai_requests"].add(1, attributes)
+                    metrics["genai_total_tokens"].add(
+                        total_tokens, attributes)
+                    metrics["genai_completion_tokens"].add(
+                        completion_tokens, attributes)
+                    metrics["genai_prompt_tokens"].add(
+                        prompt_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
+
+def start_chat_streaming_async(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.
+    """
+
+    async def wrapper(wrapped, instance, args, kwargs):
+        """
+        Wraps the 'start_chat' 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 'start_chat' method to be wrapped.
+            instance: The instance of the class where the original method is defined.
+            args: Positional arguments for the 'start_chat' method.
+            kwargs: Keyword arguments for the 'start_chat' method.
+
+        Returns:
+            The response from the original 'start_chat' method.
+        """
+
+        # Special handling for streaming response to accommodate the nature of data flow
+        async 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
+                async for event in wrapped(*args, **kwargs):
+                    llmresponse += str(event.text)
+                    yield event
+
+                # Handling exception ensure observability without disrupting operation
+                try:
+                    prompt = args[0]
+
+                    prompt_tokens = math.ceil(len(prompt) / 4)
+                    completion_tokens = math.ceil(len(llmresponse) / 4)
+                    total_tokens = prompt_tokens + completion_tokens
+
+                    model = instance._model._model_id
+
+                    # Calculate cost of the operation
+                    cost = get_chat_model_cost(model,
+                                                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_VERTEXAI)
+                    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_ENVIRONMENT,
+                                        environment)
+                    span.set_attribute(SemanticConvetion.GEN_AI_APPLICATION_NAME,
+                                        application_name)
+                    span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MODEL,
+                                        model)
+                    span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_IS_STREAM,
+                                        True)
+                    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,
+                                        total_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_VERTEXAI,
+                            SemanticConvetion.GEN_AI_ENVIRONMENT:
+                                environment,
+                            SemanticConvetion.GEN_AI_TYPE:
+                                SemanticConvetion.GEN_AI_TYPE_CHAT,
+                            SemanticConvetion.GEN_AI_REQUEST_MODEL:
+                                model
+                        }
+
+                        metrics["genai_requests"].add(1, attributes)
+                        metrics["genai_total_tokens"].add(
+                            total_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()
+
+    return wrapper
+
+def embeddings_async(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.
+    """
+
+    async def wrapper(wrapped, instance, args, kwargs):
+        """
+        Wraps the 'generate_content' 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 'generate_content' method to be wrapped.
+            instance: The instance of the class where the original method is defined.
+            args: Positional arguments for the 'generate_content' method.
+            kwargs: Keyword arguments for the 'generate_content' method.
+
+        Returns:
+            The response from the original 'generate_content' method.
+        """
+
+        with tracer.start_as_current_span(gen_ai_endpoint, kind=SpanKind.CLIENT) as span:
+            response = await wrapped(*args, **kwargs)
+
+            try:
+                prompt = args[0][0]
+
+                model = instance._model_id
+
+                prompt_tokens = int(response[0].statistics.token_count)
+
+                #Calculate cost of the operation
+                cost = get_embed_model_cost(model,
+                                            pricing_info, prompt_tokens)
+
+                # Set Span attribues
+                span.set_attribute(TELEMETRY_SDK_NAME, "openlit")
+                span.set_attribute(SemanticConvetion.GEN_AI_SYSTEM,
+                                    SemanticConvetion.GEN_AI_SYSTEM_VERTEXAI)
+                span.set_attribute(SemanticConvetion.GEN_AI_TYPE,
+                                    SemanticConvetion.GEN_AI_TYPE_EMBEDDING)
+                span.set_attribute(SemanticConvetion.GEN_AI_ENDPOINT,
+                                    gen_ai_endpoint)
+                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,
+                                    model)
+                span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_IS_STREAM,
+                                    False)
+                span.set_attribute(SemanticConvetion.GEN_AI_USAGE_PROMPT_TOKENS,
+                                    prompt_tokens)
+                span.set_attribute(SemanticConvetion.GEN_AI_USAGE_TOTAL_TOKENS,
+                                    prompt_tokens)
+                span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
+                                    cost)
+                if trace_content:
+                    span.set_attribute(SemanticConvetion.GEN_AI_CONTENT_PROMPT,
+                                        prompt)
+
+                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_VERTEXAI,
+                        SemanticConvetion.GEN_AI_ENVIRONMENT:
+                            environment,
+                        SemanticConvetion.GEN_AI_TYPE:
+                            SemanticConvetion.GEN_AI_TYPE_EMBEDDING,
+                        SemanticConvetion.GEN_AI_REQUEST_MODEL:
+                            model
+                    }
+
+                    metrics["genai_requests"].add(1, attributes)
+                    metrics["genai_total_tokens"].add(
+                        prompt_tokens, attributes)
+                    metrics["genai_prompt_tokens"].add(
+                        prompt_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