Launch Week 02 wrapped — explore all five launches
Confident AI
Docs
Pricing
Products
LLM Evaluation
LLM Evaluation

Benchmark LLM systems with research-backed metrics.

LLM Observability
LLM Observability

Trace, monitor, and alert on production LLM systems.

AI Red Teaming
AI Red Teaming

Stress-test LLM apps against adversarial attacks.

AI Governance
AI Governance

Enforce AI standards and controls across teams.

Book a DemoSign Up
Book a DemoSign Up

Toxicity

Toxicity is a single-turn safety metric to determine any toxicity in LLM's output

Overview

The toxicity metric is a single-turn safety metric that uses LLM-as-a-judge to assess whether your LLM application's output contains toxic statements.

Required Parameters

These are the parameters you must supply in your test case to run evaluations for toxicity metric:

inputstringrequired

The input you supplied to your LLM application.

actual_outputstringrequired

The final output your LLM application generates.

How Is It Calculated?

The toxicity metric uses an LLM to extract independent opinions from the actual output, then uses the same LLM to count how many of those opinions contain toxic content.


\text{Toxicity} = \frac{\text{Number of Toxic Opinions}}{\text{Total Number of Opinions}}

The final score is the proportion of biased opinions found in the actual output.

Create Locally

You can create the ToxicityMetric in deepeval as follows:

from deepeval.metrics import ToxicityMetric

metric = ToxicityMetric()

Here's a list of parameters you can configure when creating a ToxicityMetric:

thresholdnumberdefault: 0.5

A float representing the maximum passing threshold.

Unlike other metrics, the threshold for the ToxicityMetric is a maximum instead of a minimum threshold.

modelstring | Objectdefault: gpt-4.1

A string specifying which of OpenAI's GPT models to use OR any custom LLM model of type DeepEvalBaseLLM.

include_reasonbooleandefault: true

A boolean to enable the inclusion a reason for its evaluation score.

async_modebooleandefault: true

A boolean to enable concurrent execution within the measure() method.

strict_modebooleandefault: false

A boolean to enforce a binary metric score: 0 for perfection, 1 otherwise.

verbose_modebooleandefault: false

A boolean to print the intermediate steps used to calculate the metric score.

Create Remotely

For users not using deepeval python, or want to run evals remotely on Confident AI, you can use the toxicity metric by adding it to a single-turn metric collection. This will allow you to use toxicity metric for:

  • Single-turn E2E testing
  • Single-turn component-level testing
  • Online and offline evals for traces and spans
Built byConfident AI