Skip to content
pricing_optimizer.sql
Menu

Braintrust pricing

Braintrust is an active observability platform designed for AI agents, enabling teams to trace, evaluate, and discover patterns in AI agent behavior in production. It provides real-time monitoring, automated quality checks, and tools to improve AI agent performance and reliability at scale. The platform supports engineering and product teams with features like agent trace inspection, evals for quality measurement, pattern discovery, and customizable dashboards, with compliance and security certifications suitable for enterprise use.

Region
US
Source currency
USD
Audiences
Enterprise, Mid-market, and Small business

Plans and public prices

Canonical pricing source
Plan Pricing Charged by Offer type
Starter — For everyone
Free

$0 / month

 Flat · $0 / month Standalone · Free

What each plan includes

Feature names follow Braintrust’s own pricing page. Check the wording before treating features from different vendors as like-for-like.

Starter — For everyone

Core platform

Model credits

$10 credits + token rates — Monthly model credit covers AI features and built-in models without own API key

Included
Processed data

1 GB processed data + $4/GB — Included data processed per month

Included
Scores

10k scores + $2.50/1k — Scored outputs from LLM-as-a-judge, autoevals, or custom code scorers

Included
Users, projects, playgrounds, experiments, datasets

Unlimited users, projects, datasets, playgrounds, and experiments — Unlimited

Included
Human review scores

1 per project — Configurable human annotation scores

Included

Data retention and export

Data retention

14-day retention — Included data retention period

Included

How pricing changed

We check this pricing page weekly and record the changes we observe. An older date can simply mean the public page has not changed.

Observed Plans and prices Source wording Source
  • Starter — For everyone: Free
  • $0 / month
 Pricing page

Compare with alternatives