Datadog SDK

This release focuses on stability and reliability, fixing rare hangs in tracing after process forking and improving CI Visibility by resolving spurious git warnings and ensuring quarantined tests are retried.

This release focuses on stability and observability improvements, fixing bugs related to threat protection reporting, process forking hangs, and ensuring correct span handling and rendering for various Gemini response types in LLM Observability.

This release focuses on stability and correctness, fixing memory leaks in database monitoring, resolving context extraction issues in SQS data streams, and ensuring proper handling of async middleware and process forking.

This release focuses on stability, fixing a memory leak in database monitoring and resolving critical data extraction issues in SQS monitoring via botocore, alongside a Django middleware fix.

This release introduces significant enhancements to AI Guard, LLM Observability (including audio support and sampling configuration), and adds tracing support for Ray Serve and Mistral AI client calls. It also includes bug fixes and infrastructure updates like a new normalized route tag for Tornado.

This release focuses on stability, fixing an issue with thread restarts in forked children and resolving a rare crash on older CPython versions.

This release focuses on stability and correctness in tracing, particularly for async generators across various Python versions, and resolves an issue with token reporting in LLM observability for streamed OpenAI completions.

This release focuses on stability, fixing an issue with thread restarts in forked children and resolving a rare crash on older CPython versions.

This release primarily focuses on bug fixes, resolving a TypeError in async generator tracing on newer Python versions and correcting timeout handling for LLM Observability span writing.

This release focuses on stability and correctness across tracing, Celery integration, and LLM observability. Key fixes include resolving duplicate trace sending post-fork and ensuring correct metadata handling for LLM spans.

This release focuses on stability and correctness, fixing critical bugs related to tracing duplication after forking and improving LLM observability by cleaning up metadata.

This release focuses on stability improvements, specifically addressing issues related to process forking in tracing and profiling components.

This release focuses on stability improvements, specifically addressing issues related to process forking in tracing and profiling components.

This release introduces LLM Observability sampling control via the DD_LLMOBS_SAMPLE_RATE environment variable.

This release focuses on stability and data completeness, fixing several crashes related to interpreter teardown and profiling, and ensuring accurate runtime metrics collection on cgroup v2 systems.

This release focuses on stability and data integrity, fixing several bugs across LLM Observability, runtime metrics, IAST, profiling, and tracing components.

This release enhances LLM Observability by enabling reliable export of span data over APM traces and fixes several bugs across tracing, IAST, and code origin features. A change in trace API versioning occurs when LLMObs is active.

This release focuses on stability, addressing several crashes related to IAST, profiling, and pytest integration, while also improving runtime metric accuracy on cgroup v2 hosts.

This release focuses heavily on stability and memory management, fixing numerous bugs across AppSec, tracing, LLM observability, and internal components like thread handling and garbage collection. Key fixes include resolving WebSocket connection failures and memory leaks.

This release primarily focuses on stability and observability improvements, including increasing the default CI Visibility timeout, fixing APM span naming for pydantic integrations, and ensuring LLM Observability spans are not dropped due to deep JSON nesting.

This release focuses primarily on stability and performance, addressing several memory leaks related to threading, profiling, and SCA, alongside fixes for recurring timer issues.

This release focuses on stability and correctness, primarily addressing several bug fixes related to LLM Observability, tracing race conditions, and code origin inspection stability.

This release introduces significant enhancements for LLM Observability, API Security routing normalization, and improved profiling visibility on Python 3.12+. It also updates Flask instrumentation to expose more detailed resource information and enhances CI Visibility logging correlation.

This release focuses primarily on stability and performance, addressing several memory leaks related to timers, thread callbacks, native frame tracking, and SCA updates.

This release fixes critical bugs in AppSec WebSocket handling and LLM Observability streaming, and improves agentless LLMObs observability by bundling spans with APM traces.

LLMObs tracing now exports APM traces agentlessly when in agentless mode, consolidating span visibility in the UI for these users. Existing precedence rules for APM tracing remain in effect.

This release introduces significant enhancements for LLM Observability, AI Guard integration, and CI Visibility, alongside fixing several critical bugs related to coverage and tracing instrumentation. A breaking change affects how HTTP status codes are reported for Azure CosmosDB spans.

This release focuses on stability by fixing a critical tracing dependency issue related to wrapt and cleaning up unnecessary logs in LLM Observability when running agentless.

This patch resolves a bug in LLM Observability concerning Bedrock inference profiles in Langchain instrumentation, ensuring spans correctly reference the underlying model ID.

This release focuses on bug fixes across LLM Observability integrations (Bedrock, OpenAI, Claude Agent SDK) and resolves a startup failure in Starlette when tracing is enabled. It also deprecates an old tracing environment variable.

This patch release fixes a bug in telemetry reporting related to malformed distribution metadata flooding stderr with tracebacks.

This release updates Azure CosmosDB span tagging by replacing the http.status_code tag with the db.response.status_code metric, which constitutes a breaking change for users relying on the old tag.

This release introduces LLM Observability cost tagging and a configuration to exclude modules from lock profiling. It also fixes a startup deadlock related to Snowflake and resolves an issue with the lock profiler under gevent.

This release focuses on stability and correctness across several areas, including fixing memory leaks in tracing, resolving crashes in profiling, and improving telemetry accuracy for AAP. LLM Observability span hierarchy and CI Visibility upload logic have also been corrected.

This release fixes several issues related to header parsing limits for W3C tracestate and baggage, ensuring compliance with specifications and preventing unbounded work during extraction. It also resolves an import conflict affecting the Strands integration.

This release focuses on stability, fixing a memory leak in tracing spans and correcting an issue where Django async operations were incorrectly marked as errors upon cancellation.

This release introduces significant new features across AI Observability (LLM, AI Guard, ASM), CI Visibility (Bazel offline support), and various integrations like Ray, Pub/Sub, and LlamaIndex. It also includes several breaking changes related to resource naming in claude_agent_sdk and Ray tracing configuration.

This release introduces configuration for ignoring Ray actor methods, enhances AI Guard IP tagging, and fixes numerous bugs across CI Visibility, Django endpoint discovery, LLM Observability, and profiling stability.

This release introduces significant features for AI Observability (LLM Observability, AI Guard enhancements) and CI Visibility (Bazel offline support), alongside critical bug fixes and breaking changes to claude_agent_sdk span naming.

This release focuses on stability by fixing rare crashes related to internal threading, especially in forking environments like Gunicorn, and resolves data loss issues in CI Visibility when using pytest-xdist.

This release focuses on stability by fixing rare crashes related to internal threads during process forking or Python shutdown.

This release primarily addresses a rare crash issue related to internal periodic threads during forking.

This release introduces significant new features across profiling, tracing (Azure CosmosDB, MySQL async, OTLP export), and LLM Observability (LlamaIndex support, Remote Config). It also removes support for the RAGAS integration and deprecates several configuration parameters.

This release introduces significant performance improvements to profiling via Cython compilation and adds extensive new features across MLFlow, AI Guard, Azure Durable Functions, and deep enhancements to LLM Observability, including Pydantic evaluation support and incremental experiment reporting. Process tags are now propagated across many components by default.

This release addresses critical stability issues related to application forking and Python shutdown, and resolves data loss for buffered test events in CI Visibility when using pytest-xdist.

This release introduces new integrations for Azure CosmosDB, LiteLLM proxy guardrails, and LlamaIndex, alongside significant bug fixes for profiling, LLM Observability span hierarchy, and CI Visibility stability. Support for the RAGAS integration has been removed.

This release introduces new tracing support for Azure CosmosDB and llama-index, adds an AI Guardrail integration for LiteLLM, and fixes several bugs related to profiling, internal thread leaks, and LLM Observability span reporting.

This release includes bug fixes for LLM Observability span hierarchy and CI Visibility test event loss during pytest-xdist crashes. Users can enable eager flushing for the latter via an environment variable.

This release significantly enhances LLM Observability features, introduces AI Guard integration for AWS Strands Agents, and optimizes profiling performance via Cython compilation. Several components now propagate process tags for better context.

This release introduces new tracing support for Azure CosmosDB, an AI Guard integration for LiteLLM, and OTLP export capabilities. It also removes the deprecated RAGAS integration for LLM Observability.

This release focuses on stability in fork-heavy environments by fixing an internal thread leak and a post-fork crash.

This release focuses on stability by fixing bugs related to client side stats tags and addressing potential thread leaks and crashes in fork-heavy applications.

This release significantly enhances LLM Observability with new features like Pydantic evaluation support, remote evaluators, and reasoning capture. It also introduces process tag propagation across several components and updates dependencies for OpenFeature.

This release focuses on bug fixes, particularly improving resilience for CI Visibility data uploads under rate limiting and resolving issues with periodic thread execution in forked processes and stack profiling on Linux with unlimited stack size.

This release focuses on bug fixes, particularly addressing rate-limiting handling for CI Visibility data and resolving startup issues for profiling threads under specific conditions (forking or unlimited stack size).

This release introduces significant new features across LLM Observability, AI Guard integration with AWS Strands Agents, and tracing for Azure Durable Functions. It also optimizes the lock profiler via Cython compilation and propagates process tags across various monitoring payloads.

This release introduces significant performance improvements to the profiler via Cython compilation and adds extensive new features across LLM Observability, tracing integrations (MLFlow, Azure Durable Functions, Google Cloud Pub/Sub), and general telemetry propagation via process tags.

This release primarily addresses a bug in LLM Observability where the @llm decorator failed on certain return types, now logging a debug message instead. Compatibility information regarding support levels is also provided.

This release focuses on LLM Observability improvements, making size limits configurable via environment variables and fixing serialization issues for nested Pydantic models and return values in @llm decorated functions.

This release focuses on stability and correctness, fixing crashes related to Git metadata handling in CI Visibility and thread management during shutdown. It also unifies trace generation for LLM Observability spans.

This release focuses on stability and correctness, fixing several bugs related to CI Visibility, LLM Observability, thread safety during process forks, and memory profiling accuracy.

This release focuses on numerous bug fixes across CI Visibility, LLM Observability, tracing, and profiling components. Key fixes include resolving conflicts with external pytest plugins and preventing crashes during process shutdown.

This release focuses primarily on bug fixes across CI Visibility, ensuring correct behavior for pytest plugins and ITR tagging, and resolves a runtime error in the Tracing component.

This release introduces significant feature enhancements across MLFlow integration, AI Guard, LLM Observability, and adds process tag propagation across many components. Several bug fixes address issues in CI Visibility, tracing context handling, and profiler stability.

This release introduces significant enhancements to LLM Observability, including experiment status reporting, DeepEval integration, and prompt management via `LLMObs.get_prompt()`. It also includes numerous bug fixes across Profiling, AI Guard, and tracing components, alongside a deprecation warning for a future type change in `Span.parent_id`.

This release primarily focuses on bug fixes, including resolving an import-time TypeError in the profiling lock profiler related to PEP 604 type unions and fixing potential shutdown crashes.

This release addresses a critical memory corruption bug in the AAP component related to concurrent WAF calls by introducing context serialization. The estimated end-of-life date is August 2026.

This release primarily addresses a bug in LLM Observability related to token capture in the LiteLLM integration. It also provides an updated end-of-life estimate.

This release introduces inferred proxy support for Azure API Management and enables stats computation by default for Python 3.14+. Several bug fixes address tracing propagation, profiler type compatibility, and race conditions in worker threads.

This release primarily focuses on stability by adding a timeout to Unix socket connections to prevent thread I/O hangs during pre-fork shutdown.

This release focuses on stability, fixing critical memory corruption issues in the AAP component and resolving compatibility problems within CI Visibility reporting plugins.

This release introduces significant enhancements to LLM Observability, including DeepEval integration and experiment summary logging, alongside critical bug fixes for AAP memory corruption and CI Visibility reporting compatibility.

This release enhances LLM Observability by adding prompt retrieval, improved experiment span context, and dataset tagging capabilities. It also includes several bug fixes related to large dataset uploads and profiling initialization.

This release introduces significant new features for LLM Observability, including built-in evaluators and expanded provider support for LLM Judge. It also deprecates the explicit `tracer` parameter across several integrations in preparation for version 5.0.0.

This release introduces significant new features for LLM Observability, including support for Claude Agent SDK tracing and new built-in evaluators. It also includes crucial bug fixes across profiling, tracing signal handling, and dependency injection.

This release introduces significant new features for LLM Observability, including advanced evaluators and Judge capabilities, alongside support for Python 3.14 template strings in IAST. It also addresses numerous stability issues across profiling, tracing, and internal components, while deprecating the explicit `tracer` parameter in several integrations.

This release introduces significant new features for LLM Observability evaluators and IAST support for Python 3.14 template strings, alongside numerous bug fixes in profiling and tracing signal handling. Several parameters related to tracer injection are deprecated in preparation for version 5.0.0.

This release introduces significant new features for LLM Observability, IAST support for Python 3.14 template strings, and pymongo 4.12+ tracing. It also deprecates several parameters in tracing integrations in favor of the default tracer singleton and includes numerous bug fixes across profiling and tracing components.

This release introduces significant enhancements to LLM Observability with class-based evaluators and concurrent synchronous experiment execution. It also adds support for LFI detection and Tornado framework integration for AAP, alongside numerous bug fixes across profiling, exception replay, and various integrations.

This release primarily addresses a bug fix related to greenlet behavior during profiling when using `gevent.joinall`.

This release introduces significant enhancements to LLM Observability with class-based evaluators and concurrent synchronous experiment execution. It also adds support for LFI detection and Tornado integration for AAP, alongside numerous bug fixes across profiling, exception replay, and various integrations.

This release focuses on stability, addressing critical bugs in Celery integration, Data Streams Monitoring auto-enabling, and gevent profiling interactions.

This release primarily addresses a stability issue in profiling by fixing a crash related to memory profiling during forking.

This release introduces significant enhancements to LLM Observability with class-based evaluators and adds configuration for logging levels via `DD_TRACE_LOG_LEVEL`. Numerous bug fixes address issues across profiling, AWS Lambda handlers, gevent compatibility, and specific library integrations like litellm and pydantic-ai.

This release introduces new profiling capabilities for threading conditions and asyncio task relationships, enhances LLM Observability, and enables WebSocket tracing by default for ASGI. It also mandates an upgrade for the minimum supported `httpx` version to 0.25.0 and fixes several stability and correctness issues across profiling and integrations.

This release introduces significant enhancements to LLM Observability with class-based evaluators and fixes several critical bugs across AAP, aws_lambda, exception replay, litellm integration, and profiling.

This release primarily addresses a rare bug related to process hanging upon fork during profiling. It also provides an update on the estimated end-of-life date.

This release focuses on bug fixes across LLM Observability integrations (langchain, litellm) and profiling, addressing issues with span duplication, stream handling errors, and incorrect flamegraph rendering.

This release introduces significant enhancements to profiling, LLM Observability, and security analysis via AAP, while also increasing the minimum required version for the httpx dependency. Several critical bugs related to profiling stability and Django UUID tagging have been resolved.

This release focuses on bug fixes across profiling, LLM Observability, and the Anthropic integration, including resolving issues related to lock subclassing, asyncio race conditions, and multiprocessing pickling on Python 3.14+.

This release focuses on bug fixes across LLM Observability and profiling, addressing issues with trace ID propagation, lock subclassing errors, and asyncio task stack reporting.

This release primarily focuses on bug fixes, addressing an evaluation error in dynamic instrumentation and a TypeError during profiling of subclassed locks.

This release focuses on bug fixes, specifically addressing issues in LLM Observability tracing for the anthropic beta client and improving profile accuracy for off-CPU asyncio Tasks.

This release introduces significant new features for LLM Observability, enhanced profiling capabilities for asyncio locks and tasks, and business logic detection for Stripe via AAP. It also deprecates the `Hooks` class in favor of direct span utility functions.

This release introduces significant new features for LLM Observability, enhanced profiling capabilities for asyncio locks and tasks, and business logic detection for Stripe via AAP. It also deprecates the `Hooks` class in favor of direct span utility functions.