NewYour coding agent can read the release notes before it upgrades.Set up the MCP server →
PyPI · #4785 most downloaded on PyPI
AI Observability and Evaluation
Last release 4 days ago
01 Oct 2026
Ships on a steady schedule
a new release about every 8 days
Nearly every release is documented
notes for 60 of the last 60 stable releases
19 versions withdrawn
withdrawn after publishing
4 years old
700 releases · first in 2023
One column per quarter.
traces: prevent missing key exception when extracting invocation parameters in llama-index
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
add persistence for span evaluations
llm_classify (#2068) (3233d56)Remove LiteLLM model support check
Do not retry if eval was successful when using SyncExecutor
Nothing published for this version
instantiate evaluators by criteria
v1/traces HTTP endpoint to handle ExportTraceServiceRequest (3c94dea)v1/traces HTTP endpoint to handle ExportTraceServiceRequest (#1968) (3c94dea)Update llm_classify and llm_generate interfaces
llm_classify and llm_generate interfaces (#1974)llm_generate (#1965) (5999133)llm_classify and llm_generate interfaces (#1974) (9fd35a1)Nothing published for this version
Nothing published for this version
Nothing published for this version
Add LLM Tracing+Evals notebook with keyless example
openai streaming function call message support
Nothing published for this version
Nothing published for this version
Nothing published for this version
Instrument LlamaIndex streaming responses
openai streaming spans show up in the ui
### Bug Fixes * Improve rate limiter behavior
semantic conventions for tool_calls array in OpenAI ChatCompletion messages
tool_calls array in OpenAI ChatCompletion messages (#1837) (c079f00)propagate error status codes to parent spans for improved visibility into trace exceptions
evals: show span evaluations in trace details slideout
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
make the app launchable when nest_asyncio is applied
Nothing published for this version
Nothing published for this version
evals: return partial results when llm function is interrupted
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
update tracer for llama-index 0.9.0
### ⚠ BREAKING CHANGES * models: openAI 1.0 (#1716) ### Features * models: openAI 1.0
phoenix-sqlean: bump bundled sqlean to 0.28.4+10a13f9
add long-context evaluators, including map reduce and refine patterns
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Your coding agent can read these notes before it upgrades. Set up the MCP server →