genai-otel-instrument 0.1.7.dev0__tar.gz → 0.1.10.dev0__tar.gz

{genai_otel_instrument-0.1.7.dev0 → genai_otel_instrument-0.1.10.dev0}/CHANGELOG.md

@@ -6,6 +6,127 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [Unreleased]
 
+## [0.1.9.1] - 2025-01-29
+
+### Added
+
+- **Enhanced README Documentation**
+  - Added professional project logo centered at the top of README
+  - Added landing page hero image showcasing the project overview
+  - Added comprehensive Screenshots section with 5 embedded demonstration images:
+    - OpenAI instrumentation with token usage, costs, and latency metrics
+    - Ollama (local LLM) zero-code instrumentation
+    - HuggingFace Transformers with automatic token counting
+    - SmolAgents framework with complete agent workflow tracing
+    - GPU metrics collection dashboard
+  - Added links to additional screenshots (Token Cost Breakdown, OpenSearch Dashboard)
+  - Added Demo Video section with placeholder for future video content
+  - All images follow OSS documentation standards with professional formatting
+
+### Changed
+
+- **Roadmap Section Cleanup**
+  - Removed Phase 4 implementation details from roadmap (Session & User Tracking, RAG/Embedding Attributes)
+  - Phase 4 features are now fully implemented and documented in the Advanced Features section
+  - Roadmap now focuses exclusively on future releases (v0.2.0 onwards)
+
+### Improved
+
+- **Comprehensive Model Pricing Database Update**
+  - Expanded pricing coverage from 145+ to 240+ models across 15+ providers
+  - **OpenAI GPT-5 Series** (4 new models):
+    - `gpt-5` - $1.25/$10 per 1M tokens
+    - `gpt-5-2025-08-07` - $1.25/$10 per 1M tokens
+    - `gpt-5-mini` - $0.25/$2 per 1M tokens
+    - `gpt-5-nano` - $0.10/$0.40 per 1M tokens
+  - **Anthropic Claude 4/3.5 Variants** (13 new models):
+    - Claude 4 Opus series: `claude-4-opus`, `claude-opus-4`, `claude-opus-4-1`, `claude-opus-4.1` - $15/$75 per 1M tokens
+    - Claude 3.5 Sonnet: `claude-3-5-sonnet-20240620`, `claude-3-5-sonnet-20241022`, `claude-sonnet-4-5`, `claude-sonnet-4-5-20250929`, `claude-3-7-sonnet` - $3/$15 per 1M tokens
+    - Claude 3.5 Haiku: `claude-3-5-haiku-20241022` - $0.80/$4 per 1M tokens
+    - Claude Haiku 4.5: `claude-haiku-4-5` - $1/$5 per 1M tokens
+  - **XAI Grok Models** (10 new models):
+    - Grok 2: `grok-2-1212`, `grok-2-vision-1212` - $2/$10 per 1M tokens
+    - Grok 3: `grok-3` - $3/$15 per 1M tokens, `grok-3-mini` - $0.30/$0.50 per 1M tokens
+    - Grok 3 Fast: `grok-3-fast` - $5/$25 per 1M tokens, `grok-3-mini-fast` - $0.60/$4 per 1M tokens
+    - Grok 4: `grok-4` - $3/$15 per 1M tokens, `grok-4-fast` - $0.20/$0.50 per 1M tokens
+    - Image models: `grok-image`, `xai-grok-image` - $0.07 per image
+  - **Google Gemini Variants** (2 new models):
+    - `gemini-2-5-flash-image` - $0.30/$30 per 1M tokens
+    - `nano-banana` - $0.30/$30 per 1M tokens
+  - **Qwen Series** (6 new models):
+    - `qwen3-next-80b-a3b-instruct` - $0.525/$2.10 per 1M tokens
+    - `qwen3-next-80b-a3b-thinking` - $0.525/$6.30 per 1M tokens
+    - `qwen3-coder-480b-a35b-instruct` - $1/$5 per 1M tokens
+    - `qwen3-max`, `qwen-qwen3-max` - $1.20/$6 per 1M tokens
+  - **Meta Llama 4 Scout & Maverick** (6 models with updated pricing):
+    - `llama-4-scout`, `llama-4-scout-17bx16e-128k`, `meta-llama/Llama-4-Scout` - $0.15/$0.50 per 1M tokens
+    - `llama-4-maverick`, `llama-4-maverick-17bx128e-128k`, `meta-llama/Llama-4-Maverick` - $0.22/$0.85 per 1M tokens
+  - **IBM Granite Models** (13 new models):
+    - Granite 3 series: `ibm-granite-3-1-8b-instruct`, `ibm-granite-3-8b-instruct`, `granite-3-8b-instruct` - $0.20/$0.20 per 1M tokens
+    - Granite 4 series: `granite-4-0-h-small`, `granite-4-0-h-tiny`, `granite-4-0-h-micro`, `granite-4-0-micro` - $0.20/$0.20 per 1M tokens
+    - Embeddings: `granite-embedding-107m-multilingual`, `granite-embedding-278m-multilingual` - $0.10/$0.10 per 1M tokens
+    - Ollama variants: `granite:3b`, `granite:8b` - $0.20/$0.20 per 1M tokens
+  - **Mistral AI Updates** (10 new models):
+    - `mistral-large-24-11`, `mistral-large-2411` - $8/$24 per 1M tokens
+    - `mistral-small-3-1`, `mistral-small-3.1` - $1/$3 per 1M tokens
+    - `mistral-medium-3`, `mistral-medium-2025` - $0.40/$2 per 1M tokens
+    - Magistral series: `magistral-small` - $1/$3, `magistral-medium` - $3/$9 per 1M tokens
+    - Codestral: `codestral-25-01`, `codestral-2501` - $1/$3 per 1M tokens
+  - **Additional Providers**:
+    - **Sarvam AI**: `sarvam-m`, `sarvamai/sarvam-m`, `sarvam-chat` - Free (Open source)
+    - **Liquid AI**: `lfm-7b`, `liquid/lfm-7b` - $0.30/$0.60 per 1M tokens
+    - **Snowflake**: `snowflake-arctic`, `snowflake-arctic-instruct` - $0.80/$2.40 per 1M tokens, `snowflake-arctic-embed-l-v2.0` - $0.05/$0.05 per 1M tokens
+    - **NVIDIA Nemotron**: `nvidia-nemotron-4-340b-instruct` - $3/$9 per 1M tokens, `nvidia-nemotron-mini` - $0.20/$0.40 per 1M tokens, `nvidia/llama-3.1-nemotron-70b-instruct` - $0.80/$0.80 per 1M tokens
+    - **ServiceNow**: `servicenow-now-assist` - $1/$3 per 1M tokens
+  - **Pricing Corrections**:
+    - `deepseek-v3.1`: Updated to $0.56/$1.68 per 1M tokens (from $1.20/$1.20)
+    - `qwen3:3b`: Renamed to `qwen3:4b` (4B parameter model)
+  - All pricing reflects official provider rates as of October 2025
+
+## [0.1.9] - 2025-01-27
+
+### Added
+
+- **HuggingFace AutoModelForCausalLM and AutoModelForSeq2SeqLM Instrumentation**
+  - Added support for direct model usage via `AutoModelForCausalLM.generate()` and `AutoModelForSeq2SeqLM.generate()`
+  - Automatic token counting from input and output tensor shapes
+  - Cost calculation based on model parameter count (uses CostCalculator's local model pricing tiers)
+  - Span attributes: `gen_ai.system`, `gen_ai.request.model`, `gen_ai.operation.name`, token counts, costs
+  - Metrics: request counter, token counter, latency histogram, cost counter
+  - Supports generation parameters: `max_length`, `max_new_tokens`, `temperature`, `top_p`
+  - Implementation in `genai_otel/instrumentors/huggingface_instrumentor.py:184-333`
+  - Example usage in `examples/huggingface/example_automodel.py`
+  - All 443 tests pass (added 1 new test)
+
+### Fixed
+
+- **CRITICAL: Cost Tracking for OpenInference Instrumentors (smolagents, litellm, mcp)**
+  - Replaced `CostEnrichmentSpanProcessor` with `CostEnrichingSpanExporter` to properly add cost attributes
+  - **Root Cause**: SpanProcessor's `on_end()` receives immutable `ReadableSpan` objects that cannot be modified
+  - **Solution**: Custom SpanExporter that enriches span data before export, creating new ReadableSpan instances with cost attributes
+  - Cost attributes now correctly appear for smolagents, litellm, and mcp spans:
+    - `gen_ai.usage.cost.total`: Total cost in USD
+    - `gen_ai.usage.cost.prompt`: Prompt tokens cost
+    - `gen_ai.usage.cost.completion`: Completion tokens cost
+  - Supports all OpenInference semantic conventions:
+    - Model name: `llm.model_name`, `gen_ai.request.model`, `embedding.model_name`
+    - Token counts: `llm.token_count.{prompt,completion}`, `gen_ai.usage.{prompt_tokens,completion_tokens}`
+    - Span kinds: `openinference.span.kind` (LLM, EMBEDDING, CHAIN, etc.)
+  - Implementation in `genai_otel/cost_enriching_exporter.py`
+  - Updated `genai_otel/auto_instrument.py` to wrap OTLP and Console exporters
+  - Model name normalization handles provider prefixes (e.g., `openai/gpt-3.5-turbo` → `gpt-3.5-turbo`)
+  - All 442 existing tests continue to pass
+
+- **HuggingFace AutoModelForCausalLM AttributeError Fix**
+  - Fixed `AttributeError: type object 'AutoModelForCausalLM' has no attribute 'generate'`
+  - Root cause: `AutoModelForCausalLM` is a factory class; `generate()` exists on `GenerationMixin`
+  - Solution: Wrap `GenerationMixin.generate()` which all generative models inherit from
+  - This covers all model types: `AutoModelForCausalLM`, `AutoModelForSeq2SeqLM`, `GPT2LMHeadModel`, etc.
+  - Added fallback import for older transformers versions
+  - Implementation in `genai_otel/instrumentors/huggingface_instrumentor.py:184-346`
+
+## [0.1.7] - 2025-01-25
+
 ### Added
 
 - **Phase 4: Session and User Tracking (4.1)**

{genai_otel_instrument-0.1.7.dev0/genai_otel_instrument.egg-info → genai_otel_instrument-0.1.10.dev0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: genai-otel-instrument
-Version: 0.1.7.dev0
+Version: 0.1.10.dev0
 Summary: Comprehensive OpenTelemetry auto-instrumentation for LLM/GenAI applications
 Author-email: Kshitij Thakkar <kshitijthakkar@rocketmail.com>
 License: Apache-2.0
@@ -180,6 +180,12 @@ Dynamic: license-file
 
 # GenAI OpenTelemetry Auto-Instrumentation
 
+<div align="center">
+  <img src=".github/images/Logo.jpg" alt="GenAI OpenTelemetry Instrumentation Logo" width="400"/>
+</div>
+
+<br/>
+
 [![PyPI version](https://badge.fury.io/py/genai-otel-instrument.svg)](https://badge.fury.io/py/genai-otel-instrument)
 [![Python Versions](https://img.shields.io/pypi/pyversions/genai-otel-instrument.svg)](https://pypi.org/project/genai-otel-instrument/)
 [![License](https://img.shields.io/badge/License-Apache%202.0-blue.svg)](https://opensource.org/licenses/Apache-2.0)
@@ -200,6 +206,14 @@ Dynamic: license-file
 [![Semantic Conventions](https://img.shields.io/badge/OTel%20Semconv-GenAI%20v1.28-orange)](https://opentelemetry.io/docs/specs/semconv/gen-ai/)
 [![CI/CD](https://img.shields.io/badge/CI%2FCD-GitHub%20Actions-2088FF?logo=github-actions&logoColor=white)](https://github.com/Mandark-droid/genai_otel_instrument/actions)
 
+---
+
+<div align="center">
+  <img src=".github/images/Landing_Page.jpg" alt="GenAI OpenTelemetry Instrumentation Overview" width="800"/>
+</div>
+
+---
+
 Production-ready OpenTelemetry instrumentation for GenAI/LLM applications with zero-code setup.
 
 ## Features
@@ -257,7 +271,8 @@ For a more comprehensive demonstration of various LLM providers and MCP tools, r
 
 ### LLM Providers (Auto-detected)
 - **With Full Cost Tracking**: OpenAI, Anthropic, Google AI, AWS Bedrock, Azure OpenAI, Cohere, Mistral AI, Together AI, Groq, Ollama, Vertex AI
-- **Hardware/Local Pricing**: Replicate (hardware-based $/second), HuggingFace (local execution, free)
+- **Hardware/Local Pricing**: Replicate (hardware-based $/second), HuggingFace (local execution with estimated costs)
+  - **HuggingFace Support**: `pipeline()`, `AutoModelForCausalLM.generate()`, `AutoModelForSeq2SeqLM.generate()`, `InferenceClient` API calls
 - **Other Providers**: Anyscale
 
 ### Frameworks
@@ -288,6 +303,65 @@ The processor supports OpenInference semantic conventions:
 pip install genai-otel-instrument[openinference]
 ```
 
+## Screenshots
+
+See the instrumentation in action across different LLM providers and observability backends.
+
+### OpenAI Instrumentation
+Full trace capture for OpenAI API calls with token usage, costs, and latency metrics.
+
+<div align="center">
+  <img src=".github/images/Screenshots/Traces_OpenAI.png" alt="OpenAI Traces" width="900"/>
+</div>
+
+### Ollama (Local LLM) Instrumentation
+Zero-code instrumentation for local models running on Ollama with comprehensive observability.
+
+<div align="center">
+  <img src=".github/images/Screenshots/Traces_Ollama.png" alt="Ollama Traces" width="900"/>
+</div>
+
+### HuggingFace Transformers
+Direct instrumentation of HuggingFace Transformers with automatic token counting and cost estimation.
+
+<div align="center">
+  <img src=".github/images/Screenshots/Trace_HuggingFace_Transformer_Models.png" alt="HuggingFace Transformer Traces" width="900"/>
+</div>
+
+### SmolAgents Framework
+Complete agent workflow tracing with tool calls, iterations, and cost breakdown.
+
+<div align="center">
+  <img src=".github/images/Screenshots/Traces_SmolAgent_with_tool_calls.png" alt="SmolAgent Traces with Tool Calls" width="900"/>
+</div>
+
+### GPU Metrics Collection
+Real-time GPU utilization, memory, temperature, and power consumption metrics.
+
+<div align="center">
+  <img src=".github/images/Screenshots/GPU_Metrics.png" alt="GPU Metrics Dashboard" width="900"/>
+</div>
+
+### Additional Screenshots
+
+- **[Token Cost Breakdown](.github/images/Screenshots/Traces_SmolAgent_Token_Cost_breakdown.png)** - Detailed token usage and cost analysis for SmolAgent workflows
+- **[OpenSearch Dashboard](.github/images/Screenshots/GENAI_OpenSearch_output.png)** - GenAI metrics visualization in OpenSearch/Kibana
+
+---
+
+## Demo Video
+
+Watch a comprehensive walkthrough of GenAI OpenTelemetry Auto-Instrumentation in action, demonstrating setup, configuration, and real-time observability across multiple LLM providers.
+
+<div align="center">
+
+  **🎥 [Watch Demo Video](https://youtu.be/YOUR_VIDEO_ID_HERE)**
+  *(Coming Soon)*
+
+</div>
+
+---
+
 ## Cost Tracking Coverage
 
 The library includes comprehensive cost tracking with pricing data for **145+ models** across **11 providers**:
@@ -307,7 +381,10 @@ The library includes comprehensive cost tracking with pricing data for **145+ mo
 
 ### Special Pricing Models
 - **Replicate**: Hardware-based pricing ($/second of GPU/CPU time) - not token-based
-- **HuggingFace Transformers**: Local execution - no API costs
+- **HuggingFace Transformers**: Local model execution with estimated costs based on parameter count
+  - Supports `pipeline()`, `AutoModelForCausalLM.generate()`, `AutoModelForSeq2SeqLM.generate()`
+  - Cost estimation uses GPU/compute resource pricing tiers (tiny/small/medium/large)
+  - Automatic token counting from tensor shapes
 
 ### Pricing Features
 - **Differential Pricing**: Separate rates for prompt tokens vs. completion tokens
@@ -836,38 +913,6 @@ genai_otel.instrument(
 - `gen_ai.eval.bias_categories` - Detected bias types (array)
 - `gen_ai.eval.toxicity_categories` - Toxicity categories (array)
 
-#### 📊 Enhanced OpenTelemetry Compliance
-
-Completing remaining items from [OTEL_SEMANTIC_GAP_ANALYSIS_AND_IMPLEMENTATION_PLAN.md](OTEL_SEMANTIC_GAP_ANALYSIS_AND_IMPLEMENTATION_PLAN.md):
-
-**Phase 4: Optional Enhancements (✅ COMPLETED)**
-
-All Phase 4 features are now available! See the [Advanced Features](#advanced-features) section for detailed documentation.
-
-- ✅ **Session & User Tracking** - Track sessions and users across requests with custom extractor functions
-  - Configurable via `session_id_extractor` and `user_id_extractor` in `OTelConfig`
-  - Automatically adds `session.id` and `user.id` span attributes
-  - See [Session and User Tracking](#session-and-user-tracking) for usage examples
-
-- ✅ **RAG/Embedding Attributes** - Enhanced observability for retrieval-augmented generation
-  - Helper methods: `add_embedding_attributes()` and `add_retrieval_attributes()`
-  - Embedding attributes: `embedding.model_name`, `embedding.text`, `embedding.vector.dimension`
-  - Retrieval attributes: `retrieval.query`, `retrieval.document_count`, `retrieval.documents.{i}.document.*`
-  - See [RAG and Embedding Attributes](#rag-and-embedding-attributes) for usage examples
-  - Complete example: `examples/phase4_session_rag_tracking.py`
-
-**Note on Agent Workflow Tracking:**
-
-Agent workflow observability is already provided by the OpenInference Smolagents instrumentor (included when `smolagents` is in `enabled_instrumentors`). This is not a new Phase 4 feature, but an existing capability:
-
-- `openinference.span.kind: "AGENT"` - Identifies agent spans
-- `agent.name` - Agent identifier (via OpenInference)
-- `agent.iteration` - Current iteration number (via OpenInference)
-- `agent.action` - Action taken (via OpenInference)
-- `agent.observation` - Observation received (via OpenInference)
-
-Agent tracking requires Python >= 3.10 and the `smolagents` library. See [OpenInference Integration](#openinference-optional---python-310-only) for details.
-
 #### 🔄 Migration Support
 
 **Backward Compatibility:**

{genai_otel_instrument-0.1.7.dev0 → genai_otel_instrument-0.1.10.dev0}/README.md

@@ -1,5 +1,11 @@
 # GenAI OpenTelemetry Auto-Instrumentation
 
+<div align="center">
+  <img src=".github/images/Logo.jpg" alt="GenAI OpenTelemetry Instrumentation Logo" width="400"/>
+</div>
+
+<br/>
+
 [![PyPI version](https://badge.fury.io/py/genai-otel-instrument.svg)](https://badge.fury.io/py/genai-otel-instrument)
 [![Python Versions](https://img.shields.io/pypi/pyversions/genai-otel-instrument.svg)](https://pypi.org/project/genai-otel-instrument/)
 [![License](https://img.shields.io/badge/License-Apache%202.0-blue.svg)](https://opensource.org/licenses/Apache-2.0)
@@ -20,6 +26,14 @@
 [![Semantic Conventions](https://img.shields.io/badge/OTel%20Semconv-GenAI%20v1.28-orange)](https://opentelemetry.io/docs/specs/semconv/gen-ai/)
 [![CI/CD](https://img.shields.io/badge/CI%2FCD-GitHub%20Actions-2088FF?logo=github-actions&logoColor=white)](https://github.com/Mandark-droid/genai_otel_instrument/actions)
 
+---
+
+<div align="center">
+  <img src=".github/images/Landing_Page.jpg" alt="GenAI OpenTelemetry Instrumentation Overview" width="800"/>
+</div>
+
+---
+
 Production-ready OpenTelemetry instrumentation for GenAI/LLM applications with zero-code setup.
 
 ## Features
@@ -77,7 +91,8 @@ For a more comprehensive demonstration of various LLM providers and MCP tools, r
 
 ### LLM Providers (Auto-detected)
 - **With Full Cost Tracking**: OpenAI, Anthropic, Google AI, AWS Bedrock, Azure OpenAI, Cohere, Mistral AI, Together AI, Groq, Ollama, Vertex AI
-- **Hardware/Local Pricing**: Replicate (hardware-based $/second), HuggingFace (local execution, free)
+- **Hardware/Local Pricing**: Replicate (hardware-based $/second), HuggingFace (local execution with estimated costs)
+  - **HuggingFace Support**: `pipeline()`, `AutoModelForCausalLM.generate()`, `AutoModelForSeq2SeqLM.generate()`, `InferenceClient` API calls
 - **Other Providers**: Anyscale
 
 ### Frameworks
@@ -108,6 +123,65 @@ The processor supports OpenInference semantic conventions:
 pip install genai-otel-instrument[openinference]
 ```
 
+## Screenshots
+
+See the instrumentation in action across different LLM providers and observability backends.
+
+### OpenAI Instrumentation
+Full trace capture for OpenAI API calls with token usage, costs, and latency metrics.
+
+<div align="center">
+  <img src=".github/images/Screenshots/Traces_OpenAI.png" alt="OpenAI Traces" width="900"/>
+</div>
+
+### Ollama (Local LLM) Instrumentation
+Zero-code instrumentation for local models running on Ollama with comprehensive observability.
+
+<div align="center">
+  <img src=".github/images/Screenshots/Traces_Ollama.png" alt="Ollama Traces" width="900"/>
+</div>
+
+### HuggingFace Transformers
+Direct instrumentation of HuggingFace Transformers with automatic token counting and cost estimation.
+
+<div align="center">
+  <img src=".github/images/Screenshots/Trace_HuggingFace_Transformer_Models.png" alt="HuggingFace Transformer Traces" width="900"/>
+</div>
+
+### SmolAgents Framework
+Complete agent workflow tracing with tool calls, iterations, and cost breakdown.
+
+<div align="center">
+  <img src=".github/images/Screenshots/Traces_SmolAgent_with_tool_calls.png" alt="SmolAgent Traces with Tool Calls" width="900"/>
+</div>
+
+### GPU Metrics Collection
+Real-time GPU utilization, memory, temperature, and power consumption metrics.
+
+<div align="center">
+  <img src=".github/images/Screenshots/GPU_Metrics.png" alt="GPU Metrics Dashboard" width="900"/>
+</div>
+
+### Additional Screenshots
+
+- **[Token Cost Breakdown](.github/images/Screenshots/Traces_SmolAgent_Token_Cost_breakdown.png)** - Detailed token usage and cost analysis for SmolAgent workflows
+- **[OpenSearch Dashboard](.github/images/Screenshots/GENAI_OpenSearch_output.png)** - GenAI metrics visualization in OpenSearch/Kibana
+
+---
+
+## Demo Video
+
+Watch a comprehensive walkthrough of GenAI OpenTelemetry Auto-Instrumentation in action, demonstrating setup, configuration, and real-time observability across multiple LLM providers.
+
+<div align="center">
+
+  **🎥 [Watch Demo Video](https://youtu.be/YOUR_VIDEO_ID_HERE)**
+  *(Coming Soon)*
+
+</div>
+
+---
+
 ## Cost Tracking Coverage
 
 The library includes comprehensive cost tracking with pricing data for **145+ models** across **11 providers**:
@@ -127,7 +201,10 @@ The library includes comprehensive cost tracking with pricing data for **145+ mo
 
 ### Special Pricing Models
 - **Replicate**: Hardware-based pricing ($/second of GPU/CPU time) - not token-based
-- **HuggingFace Transformers**: Local execution - no API costs
+- **HuggingFace Transformers**: Local model execution with estimated costs based on parameter count
+  - Supports `pipeline()`, `AutoModelForCausalLM.generate()`, `AutoModelForSeq2SeqLM.generate()`
+  - Cost estimation uses GPU/compute resource pricing tiers (tiny/small/medium/large)
+  - Automatic token counting from tensor shapes
 
 ### Pricing Features
 - **Differential Pricing**: Separate rates for prompt tokens vs. completion tokens
@@ -656,38 +733,6 @@ genai_otel.instrument(
 - `gen_ai.eval.bias_categories` - Detected bias types (array)
 - `gen_ai.eval.toxicity_categories` - Toxicity categories (array)
 
-#### 📊 Enhanced OpenTelemetry Compliance
-
-Completing remaining items from [OTEL_SEMANTIC_GAP_ANALYSIS_AND_IMPLEMENTATION_PLAN.md](OTEL_SEMANTIC_GAP_ANALYSIS_AND_IMPLEMENTATION_PLAN.md):
-
-**Phase 4: Optional Enhancements (✅ COMPLETED)**
-
-All Phase 4 features are now available! See the [Advanced Features](#advanced-features) section for detailed documentation.
-
-- ✅ **Session & User Tracking** - Track sessions and users across requests with custom extractor functions
-  - Configurable via `session_id_extractor` and `user_id_extractor` in `OTelConfig`
-  - Automatically adds `session.id` and `user.id` span attributes
-  - See [Session and User Tracking](#session-and-user-tracking) for usage examples
-
-- ✅ **RAG/Embedding Attributes** - Enhanced observability for retrieval-augmented generation
-  - Helper methods: `add_embedding_attributes()` and `add_retrieval_attributes()`
-  - Embedding attributes: `embedding.model_name`, `embedding.text`, `embedding.vector.dimension`
-  - Retrieval attributes: `retrieval.query`, `retrieval.document_count`, `retrieval.documents.{i}.document.*`
-  - See [RAG and Embedding Attributes](#rag-and-embedding-attributes) for usage examples
-  - Complete example: `examples/phase4_session_rag_tracking.py`
-
-**Note on Agent Workflow Tracking:**
-
-Agent workflow observability is already provided by the OpenInference Smolagents instrumentor (included when `smolagents` is in `enabled_instrumentors`). This is not a new Phase 4 feature, but an existing capability:
-
-- `openinference.span.kind: "AGENT"` - Identifies agent spans
-- `agent.name` - Agent identifier (via OpenInference)
-- `agent.iteration` - Current iteration number (via OpenInference)
-- `agent.action` - Action taken (via OpenInference)
-- `agent.observation` - Observation received (via OpenInference)
-
-Agent tracking requires Python >= 3.10 and the `smolagents` library. See [OpenInference Integration](#openinference-optional---python-310-only) for details.
-
 #### 🔄 Migration Support
 
 **Backward Compatibility:**

{genai_otel_instrument-0.1.7.dev0 → genai_otel_instrument-0.1.10.dev0}/examples/anthropic/example.py

@@ -15,7 +15,7 @@ client = Anthropic()
 
 # Make a simple message request
 message = client.messages.create(
-    model="claude-3-5-sonnet-20241022",
+    model="claude-haiku-4-5",
     max_tokens=150,
     messages=[{"role": "user", "content": "Explain OpenTelemetry in one sentence."}],
 )

genai_otel_instrument-0.1.10.dev0/examples/huggingface/example_automodel.py

@@ -0,0 +1,89 @@
+"""HuggingFace AutoModelForCausalLM Example with Token Counting and Cost Tracking.
+
+This example demonstrates:
+1. Auto-instrumentation of AutoModelForCausalLM.generate()
+2. Automatic token counting (prompt + completion tokens)
+3. Cost calculation for local model inference
+4. Full observability with traces and metrics
+
+Requirements:
+    pip install transformers torch
+"""
+
+import genai_otel
+
+# Auto-instrument HuggingFace Transformers
+genai_otel.instrument()
+
+from transformers import AutoModelForCausalLM, AutoTokenizer
+
+print("\n" + "=" * 80)
+print("Loading model and tokenizer...")
+print("=" * 80 + "\n")
+
+# Load a small model for testing (117M parameters)
+model_name = "Qwen/Qwen3-0.6B"
+tokenizer = AutoTokenizer.from_pretrained(model_name)
+model = AutoModelForCausalLM.from_pretrained(model_name)
+
+print(f"Model loaded: {model_name}")
+print(f"Model config: {model.config._name_or_path}\n")
+
+# Prepare input
+prompt = "The future of AI is"
+inputs = tokenizer(prompt, return_tensors="pt")
+
+print(f"Prompt: '{prompt}'")
+print(f"Input tokens: {inputs['input_ids'].shape[-1]}\n")
+
+print("=" * 80)
+print("Generating text (instrumented)...")
+print("=" * 80 + "\n")
+
+# Generate text - This is automatically instrumented!
+# The wrapper will:
+# - Create a span with model info
+# - Count input tokens (from input_ids.shape)
+# - Count output tokens (from generated sequence length)
+# - Calculate cost based on GPT-2's parameter count (117M -> tier pricing)
+# - Record metrics for tokens and cost
+outputs = model.generate(
+    inputs["input_ids"],
+    max_new_tokens=50,
+    temperature=0.7,
+    do_sample=True,
+    pad_token_id=tokenizer.eos_token_id,
+)
+
+# Decode the generated text
+generated_text = tokenizer.decode(outputs[0], skip_special_tokens=True)
+
+print(f"Generated text: {generated_text}\n")
+print(f"Total output tokens: {outputs.shape[-1]}")
+print(f"Input tokens: {inputs['input_ids'].shape[-1]}")
+print(f"Generated (new) tokens: {outputs.shape[-1] - inputs['input_ids'].shape[-1]}\n")
+
+print("=" * 80)
+print("Telemetry captured:")
+print("=" * 80)
+print("✓ Span created: huggingface.model.generate")
+print("✓ Attributes set:")
+print(f"  - gen_ai.system: huggingface")
+print(f"  - gen_ai.request.model: {model_name}")
+print(f"  - gen_ai.operation.name: text_generation")
+print(f"  - gen_ai.usage.prompt_tokens: {inputs['input_ids'].shape[-1]}")
+print(f"  - gen_ai.usage.completion_tokens: {outputs.shape[-1] - inputs['input_ids'].shape[-1]}")
+print(f"  - gen_ai.usage.total_tokens: {outputs.shape[-1]}")
+print("  - gen_ai.usage.cost.total: $X.XXXXXX (estimated)")
+print("  - gen_ai.usage.cost.prompt: $X.XXXXXX")
+print("  - gen_ai.usage.cost.completion: $X.XXXXXX")
+print("\n✓ Metrics recorded:")
+print("  - gen_ai.requests counter")
+print("  - gen_ai.client.token.usage (prompt + completion)")
+print("  - gen_ai.client.operation.duration histogram")
+print("  - gen_ai.usage.cost counter")
+print("\n✓ Traces and metrics exported to OTLP endpoint!")
+print("=" * 80)
+
+print("\nNote: Cost is estimated based on model size (GPT-2 = 117M params)")
+print("Local models are free to run, but costs reflect GPU/compute resources.")

{genai_otel_instrument-0.1.7.dev0 → genai_otel_instrument-0.1.10.dev0}/genai_otel/__version__.py

@@ -28,7 +28,7 @@ version_tuple: VERSION_TUPLE
 commit_id: COMMIT_ID
 __commit_id__: COMMIT_ID
 
-__version__ = version = '0.1.7.dev0'
-__version_tuple__ = version_tuple = (0, 1, 7, 'dev0')
+__version__ = version = '0.1.10.dev0'
+__version_tuple__ = version_tuple = (0, 1, 10, 'dev0')
 
-__commit_id__ = commit_id = 'g6e54041fe'
+__commit_id__ = commit_id = 'gb29b5b98f'

{genai_otel_instrument-0.1.7.dev0 → genai_otel_instrument-0.1.10.dev0}/genai_otel/auto_instrument.py

@@ -19,6 +19,7 @@ from opentelemetry.sdk.trace.export import BatchSpanProcessor, ConsoleSpanExport
 from .config import OTelConfig
 from .cost_calculator import CostCalculator
 from .cost_enrichment_processor import CostEnrichmentSpanProcessor
+from .cost_enriching_exporter import CostEnrichingSpanExporter
 from .gpu_metrics import GPUMetricsCollector
 from .mcp_instrumentors import MCPInstrumentorManager
 from .metrics import (
@@ -169,14 +170,17 @@ def setup_auto_instrumentation(config: OTelConfig):
 
     set_global_textmap(TraceContextTextMapPropagator())
 
-    # Add cost enrichment processor for OpenInference instrumentors
-    # This enriches spans from smolagents, litellm, mcp with cost attributes
+    # Add cost enrichment processor for custom instrumentors (OpenAI, Ollama, etc.)
+    # These instrumentors set cost attributes directly, so processor is mainly for logging
+    # Also attempts to enrich OpenInference spans (smolagents, litellm, mcp), though
+    # the processor can't modify ReadableSpan - the exporter below handles that
+    cost_calculator = None
     if config.enable_cost_tracking:
         try:
             cost_calculator = CostCalculator()
             cost_processor = CostEnrichmentSpanProcessor(cost_calculator)
             tracer_provider.add_span_processor(cost_processor)
-            logger.info("Cost enrichment processor added for OpenInference instrumentors")
+            logger.info("Cost enrichment processor added")
         except Exception as e:
             logger.warning(f"Failed to add cost enrichment processor: {e}", exc_info=True)