Scoring API

from withpi import PiClient

pi = PiClient()
scores = pi.scoring_system.score(
  llm_input="Pi Labs",
  llm_output="Score anything with Pi Labs today!",
  scoring_spec=[{"question": "Is there a strong call to action?"}]
)
print(scores.total_score)

Start Scoring Build a scorer

Lightweight and powerful.

https://api.withpi.ai

Use any language

Our REST-API is openapi compatible, and provides a universal way to access our models

const spec = [{
  question: "My first question",
}, {
  question: "My second question",
}]

Simple interface

Scoring specs are a simple JSON payload that can be through constructed our copilot, or your own systems.

200ms

Minimal overhead

Our models can judge criteria 500x more efficiently than LLM as a judge, unlocking online use cases for scoring.

Example Applications

Scoring

Bring some questions and a response to score, and see how Pi handles them.

Model Comparison

Confidently compare model outputs to see which one hits the price/performance

Dataset Filtering

Filter large amounts of data and arrive at a better set for training or evaluation.

Generate Synthetic Training Data

Generate a synthetic training set and filter it against your scoring system.

Calibration

Tune your scoring system against user feedback and rating.

Ready to start scoring? Grab a free API key to get started. Don't hesitate to reach out for enterprise needs.

Create an API Key Talk to an expert

Explore Pi Enterprise and interact with a dedicated quality engineer to help you integrate.

Explore Enterprise

Home Docs Pricing Support