OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
AI Evaluation & Observability

Parea SDK for Python

Parea SDK for Python is a Python package for evaluating and testing LLM applications, instrumenting logs and traces for monitoring, and supporting collaborative prompt deployment.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Parea SDK for Python — OpenViST project profile in AI Evaluation & Observability.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use it when an LLM application already has a Python workflow and you want dataset-based experiments or code-level logging and tracing.

Who it is for

  • Python developers building LLM-powered applications.
  • Teams that need to score application behavior against datasets.
  • Engineers who want traces or model-call logging in an LLM application.
  • Engineers and subject-matter experts collaborating on prompts.

Documented uses

  • Run an application function against example data and score its outputs.
  • Review experiment outputs, scores and traces.
  • Capture hierarchical traces across application steps.
  • Log calls made through wrapped OpenAI or Anthropic clients.
Before you choose

Requirements & limits

The documented workflows initialize the SDK with a Parea API key. The repository specifies an Apache-2.0 license.

  • The documented installation options are pip and Poetry.
  • Experiment evaluation is defined around a function, dataset and evaluation function.
  • Client wrapping is shown for OpenAI and Anthropic; hierarchical traces use the trace decorator.
  • The README lists integrations including LangChain, Instructor, DSPy and LiteLLM.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Run a dataset-based LLM evaluation

Score a Python function over example data and create an experiment report.

Before you start

  • Python environment with the parea-ai package installed using pip or Poetry.
  • Parea API key.
  • A function to evaluate and a dataset of examples.

Steps

  1. Install the package with the command below or poetry add parea-ai.

    Documented command
    pip install -U parea-ai

  2. Initialize Parea with your API key and import trace and an evaluation function such as levenshtein.

  3. Decorate the function with @trace(eval_funcs=[levenshtein]) and define dataset examples.

  4. Pass the function and dataset to p.experiment(...) and call .run().

Expected outcome

An experiment report containing outputs, scores and traces for dataset samples.

Guide limitations

  • The README's example uses a greeting function and Levenshtein scoring; it does not establish which evaluation metric is appropriate for other tasks.
Documentation 1 ↗
Source-based practical guide

Instrument OpenAI calls for logging

Wrap an OpenAI client so its calls are logged through Parea.

Before you start

  • Python environment with the parea-ai package and OpenAI client available.
  • OpenAI API key and Parea API key.

Steps

  1. Create an OpenAI client using its API key.

  2. Initialize Parea with the Parea API key.

  3. Call p.wrap_openai_client(client) before making model calls through that client.

Expected outcome

The README describes automatic logging of OpenAI calls made through the wrapped client.

Guide limitations

  • The example covers OpenAI client wrapping; it does not specify retention, log controls or other service settings.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
17 October 2026
Original website title
parea-ai/parea-sdk-py · Official homepage title · retained in its source language
How we review sourcesSuggest a correction