NewYour coding agent can read the release notes before it upgrades.Set up the MCP server →
PyPI · #840 most downloaded on PyPI
Langfuse Python SDK - LLM observability/tracing, datasets, experiments, LLM-as-a-judge evaluation, and prompt management
Last release 2 days ago
16 Sep 2026
Ships on a steady schedule
a new release about every 2 weeks
Nearly every release is documented
notes for 60 of the last 60 stable releases
3 versions withdrawn
withdrawn after publishing
3 years old
551 releases · first in 2023
One column per quarter.
fix(langchain): anthropic usage parsing by @hassiebp in https://github.com/langfuse/langfuse-python/pull/1194
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.60.5...v2.60.6
fix(openai): usage parsing for streamed responses by @hassiebp in https://github.com/langfuse/langfuse-python/pull/1184
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.60.4...v2.60.5
feat: add env filter to fetch_traces() and fetch_observations() by @marcklingen in https://github.com/langfuse/langfuse-python/pull/1166
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.60.3...v2.60.4
chore(deps): bump llama-index from 0.12.5 to 0.12.9 by @dependabot in https://github.com/langfuse/langfuse-python/pull/1149
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.60.2...v2.60.3
fix(cost-tracking): handle none values in OpenAI schema by @hassiebp in https://github.com/langfuse/langfuse-python/pull/1152
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.60.1...v2.60.2
chore: fix typo in function name by @KimHyeonwoo in https://github.com/langfuse/langfuse-python/pull/1147
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.60.0...v2.60.1
feat(openai): add Response API support by @hassiebp in https://github.com/langfuse/langfuse-python/pull/1141
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.59.7...v2.60.0
feat(client): add native environments by @hassiebp in https://github.com/langfuse/langfuse-python/pull/1122
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.59.6...v2.59.7
fix(openai): handle missing text property on streamed completion by @hassiebp in https://github.com/langfuse/langfuse-python/pull/1126
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.59.5...v2.59.6
Resolve runtime error with openai extension when metadata is missing by @seratch in https://github.com/langfuse/langfuse-python/pull/1115
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.59.4...v2.59.5
fix(langchain): cached token usage by @hassiebp in https://github.com/langfuse/langfuse-python/pull/1121
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.59.3...v2.59.4
fix(openai): implement aclose on async stream responses by @hassiebp in https://github.com/langfuse/langfuse-python/pull/1112
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.59.2...v2.59.3
fix(serializer): NaN handling by @hassiebp in https://github.com/langfuse/langfuse-python/pull/1107
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.59.1...v2.59.2
perf(ingestion): make max event and batch size configurable by @hassiebp in https://github.com/langfuse/langfuse-python/pull/1103
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.59.0...v2.59.1
feat(api): expose public api client by @hassiebp in https://github.com/langfuse/langfuse-python/pull/1098
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.58.2...v2.59.0
fix(openai): handle usage object without mutation by @hassiebp in https://github.com/langfuse/langfuse-python/pull/1101
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.58.1...v2.58.2
fix(llama-index): workflow integration (#1095) by @petros94 in https://github.com/langfuse/langfuse-python/pull/1097
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.58.0...v2.58.1
feat: update prompts and invalidate cache by @maxdeichmann in https://github.com/langfuse/langfuse-python/pull/1082
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.57.13...v2.58.0
fix(langchain): add run_id check on_llm_error by @hassiebp in https://github.com/langfuse/langfuse-python/pull/1089
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.57.12...v2.57.13
Nothing published for this version
fix(decorators): merge trace tags and metadata on updates by @hassiebp in https://github.com/langfuse/langfuse-python/commit/f0ab24d5accf217c59df61f84
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.57.11...v2.57.12
fix(openai): allow non in token_details by @hassiebp in https://github.com/langfuse/langfuse-python/commit/9a694db43147e801193018124c282c88c9f5669d
fix(langchain): mitigate missing run_id on_chain_error by @hassiebp in https://github.com/langfuse/langfuse-python/pull/1078
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.57.9...v2.57.10
chore(deps): bump django from 5.0.10 to 5.0.11 in /examples/django_example by @dependabot in https://github.com/langfuse/langfuse-python/pull/1075
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.57.8...v2.57.9
perf(decorator): remove get_caller_module by @hassiebp in https://github.com/langfuse/langfuse-python/pull/1074
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.57.7...v2.57.8
fix(api-models): usage unit to optional by @hassiebp in https://github.com/langfuse/langfuse-python/pull/1073
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.57.6...v2.57.7
chore(deps-dev): bump virtualenv from 20.26.1 to 20.26.6 by @dependabot in https://github.com/langfuse/langfuse-python/pull/1070
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.57.5...v2.57.6
fix(serializer): remove unnecessary log statement by @Kosmasu and @hassiebp in https://github.com/langfuse/langfuse-python/pull/1061
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.57.4...v2.57.5
fix: add project_id to urls returned by langfuse_context.get_current_trace_url by @marcklingen in https://github.com/langfuse/langfuse-python/pull/105
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.57.3...v2.57.4
fix(dataset+decorator): item observer with top-level generation by @hassiebp in https://github.com/langfuse/langfuse-python/pull/1055
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.57.2...v2.57.3
perf: memory usage on repeat instantiation
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.57.1...v2.57.2
chore(singleton): shutdown on reset by @hassiebp in https://github.com/langfuse/langfuse-python/pull/1050
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.57.0...v2.57.1
Nothing published for this version
feat(media): utility to resolve media references with media content by @hassiebp in https://github.com/langfuse/langfuse-python/pull/1036
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.56.1...v2.56.2
fix(client): pass timeout to FernLangfuse by @hassiebp in https://github.com/langfuse/langfuse-python/pull/1030
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.56.0...v2.56.1
chore(deps): bump the langchain group across 1 directory with 2 updates by @dependabot in https://github.com/langfuse/langfuse-python/pull/1014
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.55.0...v2.56.0
chore: remove Python 3.8 support due to EOL
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.54.1...v2.55.0
fix(media): allow setting IO media via decorator update by @hassiebp in https://github.com/langfuse/langfuse-python/pull/1005
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.54.0...v2.54.1
feat(core): add multimodal support by @hassiebp in https://github.com/langfuse/langfuse-python/pull/996
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.53.9...v2.54.0
Nothing published for this version
perf: move serialization to background threads by @hassiebp in https://github.com/langfuse/langfuse-python/pull/994
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.53.8...v2.53.9
fix(datasets): encoding by @hassiebp in https://github.com/langfuse/langfuse-python/pull/993
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.53.7...v2.53.8
fix(openai): revert default stream option setting by @hassiebp in https://github.com/langfuse/langfuse-python/pull/992
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.53.6...v2.53.7
fix(serializer): reduce log level to debug on failed serialization by @hassiebp in https://github.com/langfuse/langfuse-python/pull/991
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.53.5...v2.53.6
fix(serializer): pydantic compat v1 v2 by @hassiebp in https://github.com/langfuse/langfuse-python/pull/988
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.53.4...v2.53.5
feat(openai): parse usage if stream_options has include_usage by @hassiebp in https://github.com/langfuse/langfuse-python/pull/987
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.53.3...v2.53.4
fix(datasets): url encode dataset name and run name by @hassiebp in https://github.com/langfuse/langfuse-python/pull/980
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.53.2...v2.53.3
fix(llama-index): CompletionResponse Serialization by @hassiebp
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.53.1...v2.53.2
fix: 'NoneType' object has no attribute '__dict__' by @bettlebrox in https://github.com/langfuse/langfuse-python/pull/978
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.53.0...v2.53.1
feat(client): allow masking event input and output by @shawnzhu and @hassiebp in https://github.com/langfuse/langfuse-python/pull/977
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.52.2...v2.53.0
docs(llama-index): deprecate CallbackHandler and promote Instrumentor by @hassiebp in https://github.com/langfuse/langfuse-python/pull/971
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.52.2...v2.52.2
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.52.1...v2.52.2
fix(decorators): stack trace on failed auth_check by @hassiebp in https://github.com/langfuse/langfuse-python/pull/961
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.52.0...v2.52.1
feat(llama-index): add LlamaIndexInstrumentor for use with instrumentation module instead of LlamaIndexSpanHandler by @hassiebp in https://github.com/
fix(openai): structured output parsing with openai >= 1.50 by @hassiebp in https://github.com/langfuse/langfuse-python/pull/957
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.51.4...v2.51.5
fix(serializer): Fix None serialization without langchain installed by @abrasseu in https://github.com/langfuse/langfuse-python/pull/955
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.51.3...v2.51.4
fix(langchain): time to first token by @mrdrprofuroboros and @hassiebp in https://github.com/langfuse/langfuse-python/pull/944
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.51.2...v2.51.3
fix(prompts): remove traceback from fetch error logs by @hassiebp in https://github.com/langfuse/langfuse-python/pull/940
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.51.1...v2.51.2
fix(langchain): use ls_model_name for parsing from @hassiebp
feat(langchain): time to first token on streamed generations by @hassiebp in https://github.com/langfuse/langfuse-python/pull/938
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.50.3...v2.51.0
fix(serializer): support numpy scalars by @hassiebp in https://github.com/langfuse/langfuse-python/pull/936
Full Changelog: https://github.com/langfuse/langfuse-python/compare/v2.50.2...v2.50.3
Your coding agent can read these notes before it upgrades. Set up the MCP server →