0.45.1

v0.45.1 (2025-08-13)

Fix

• langchain: context detach exception (#3255)
• mcp: MCP Instrumentation: streamablehttp_client Parameter Corruption (#3199)

[main f706a9f] bump: version 0.45.0 → 0.45.1
58 files changed, 64 insertions(+), 57 deletions(-)

0.45.0

v0.45.0 (2025-08-12)

Feat

• datasets: add dataset and datasets functionality (#3247)

Fix

• anthropic: support with_raw_response wrapper for span generation (#3250)
• langchain: fix nesting of langgraph spans (#3206)
• langchain: Add "dont_throw" to "on_llm_end" and remove blank file (#3232)

[main 2708764] bump: version 0.44.3 → 0.45.0
58 files changed, 69 insertions(+), 57 deletions(-)

0.44.3

v0.44.3 (2025-08-12)

Fix

• sdk: avoid initializing metrics exporter on custom tracing config (#3249)
• openai: propagate span IDs properly to events (#3243)

[main 2bb91eb] bump: version 0.44.2 → 0.44.3
58 files changed, 64 insertions(+), 57 deletions(-)

0.44.2

v0.44.2 (2025-08-11)

Fix

• openai: dynamically import types for 1.99 (#3244)
• langchain: Added new method for fetching model name from association metadata (#3237)

[main 3926524] bump: version 0.44.1 → 0.44.2
58 files changed, 64 insertions(+), 57 deletions(-)

0.44.1

v0.44.1 (2025-08-04)

Fix

• mcp: do not override meta pydantic types (#3179)

[main 8ee725c] bump: version 0.44.0 → 0.44.1
58 files changed, 63 insertions(+), 57 deletions(-)

0.44.0

v0.44.0 (2025-08-03)

Feat

• sdk: support multiple span processors (#3207)
• semantic-conentions-ai: add LLMVendor enum to semantic conventions (#3170)

Fix

• langchain: spans dictionary memory leak (#3216)
• openai-agents: use framework's context to infer trace (#3215)
• sdk: respect truncation otel environment variable (#3212)
• anthropic: async stream manager (#3220)
• langchain: populate metadata as span attributes in batch operations (#3218)
• anthropic: various fixes around tools parsing (#3204)
• qdrant: fix qdrant-client auto instrumentation condition (#3208)
• instrumentation: remove param enrich_token_usage and simplify token calculation (#3205)
• langchain: ensure llm spans are created for sync cases (#3201)
• openai: support for openai non-consumed streams (#3155)

[main cfe309d] bump: version 0.43.1 → 0.44.0
58 files changed, 77 insertions(+), 57 deletions(-)

0.43.1

v0.43.1 (2025-07-23)

Fix

• langchain: added vendors to llm calls (#3165)

[main 87f8932] bump: version 0.43.0 → 0.43.1
58 files changed, 63 insertions(+), 57 deletions(-)

0.43.0

v0.43.0 (2025-07-22)

Feat

• prompts: add tool function support (#3153)

Fix

• llamaindex: structured llm model and temperature parsing (#3159)
• langchain: report token usage histogram (#3059)
• openai: prioritize api-provided token over tiktoken calculation (#3142)
• milvus: Add metrics support (#3013)

[main 6704f15] bump: version 0.42.0 → 0.43.0
58 files changed, 70 insertions(+), 57 deletions(-)

0.42.0

v0.42.0 (2025-07-17)

Feat

• llamaindex: support llamaparse instrumentation (#3103)
• milvus: add semantic convention for Milvus DB metrics (#3015)

Fix

• openai-agents: fix broken traces with agents handoff on run_stream (#3143)
• traceloop-sdk: redefine histogram bucket boundaries (#3129)

[main 8aaec30] bump: version 0.41.0 → 0.42.0
58 files changed, 69 insertions(+), 57 deletions(-)

0.41.0

v0.41.0 (2025-07-13)

Feat

• openai-agents: initial instrumentation; collect OpenAI agent traces and metrics (#2966)
• google-generativeai: implement emitting events in addition to current behavior (#2887)
• vertexai: implement emitting events in addition to current behavior (#2942)
• langchain: implement emitting events in addition to current behavior (#2889)
• anthropic: implement emitting events in addition to current behavior (#2884)
• bedrock: implement emitting events in addition to current behavior (#2885)
• llamaindex: implement emitting events in addition to current behavior (#2941)
• watsonx: implement emitting events in addition to current behavior (#2896)
• cohere: implement emitting events in addition to current behavior (#2886)
• groq: implement emitting events in addition to current behavior (#2888)
• sagemaker: implement emitting events in addition to current behavior (#2894)
• together: implement emitting events in addition to current behavior (#2895)
• replicate: implement emitting events in addition to current behavior (#2893)
• ollama: implement emitting events in addition to current behavior (#2891)
• mistralai: implement emitting events in addition to current behavior (#2890)
• vendor matching (#3062)
• semconv: add an attribute for output schema (#3064)
• transformers: implement the support to emitting events in addition to current behavior (#2940)
• alephalpha: implement emitting events in addition to current behavior (#2880)
• mcp: Add support for mcp streamable http transport type (#3049)
• milvus: Add error.type attribute from OpenTelemetry Semantic Conventions (#3009)
• openai: OpenAI responses minimal instrumentation (#3052)
• ollama: add meter STTG to ollama instrumentation (#3053)
• openai: implement emitting events in addition to current behavior (#2892)

Fix

• align semconv deps (#3106)
• sagemaker: add should_send_prompts checks (#3072)
• watsonx: add should_send_prompts check to model response (#3071)
• groq: add should_send_prompts checks (#3074)
• ollama: add should_send_prompts check (#3073)
• transformers: add should_send_prompts checks (#3070)
• groq: wrong system attribute was given (#3069)
• openai: record exception as span events as well (#3067)
• openai: add request schema attribute (#3065)
• mcp: add support for error_type in mcp instrumentation (#3050)
• google gemini insturmentation (#3055)
• openai: completions.parse out of beta, azure remove double-slash (#3051)

[main 6454263] bump: version 0.40.14 → 0.41.0
58 files changed, 101 insertions(+), 57 deletions(-)