How Good Is Your AI Agent?

Free benchmark tool for AI agents. Get a score from 1-100, detailed recommendations, and compare on the public leaderboard. Works with LangChain, AutoGPT, CrewAI, Claude Code, or any framework.

Run Free Benchmark View Leaderboard
Windows macOS Linux

Your data never stored • No account needed • Results in 60 seconds

bash 
$ sudodog-benchmark

   _____ __  ______  ____  ____  ____  ______
  / ___// / / / __ \/ __ \/ __ \/ __ \/ ____/
  \__ \/ / / / / / / / / / / / / / / / / __  
 ___/ / /_/ / /_/ / /_/ / /_/ / /_/ / /_/ /  
/____/\____/_____/\____/\____/\____/\____/   

Scanning for AI agents...
Found: langchain (PID 12847)

Monitoring agent behavior (30s)...
[████████████████████████████] 100%

Analyzing with AI...

═══════════════════════════════════════
  BENCHMARK RESULTS
═══════════════════════════════════════
  Score: 78/100 (B+)
  
  + Good error handling
  + Efficient API usage
  - High memory usage detected
  - Consider adding rate limiting
  
  Full Report: https://sudodog.com/report/abc123
  Leaderboard: https://sudodog.com/leaderboard

How It Works

Three simple steps. No account required. Free forever.

1. Install

One command. Works on Linux, macOS, and Windows.

pip install sudodog

2. Run Benchmark

Start your agent, then run the benchmark. It auto-detects running agents.

sudodog-benchmark

3. Get Your Score

AI-powered analysis gives you a score, grade, and actionable recommendations.

78/100 B+

What You Get

Score & Grade

Get a score from 1-100 with a letter grade (A+ to F). Know exactly where your agent stands.

Detailed Report

See what's working well and what needs improvement. Shareable link for your team.

AI Recommendations

AI-powered analysis provides specific, actionable fixes you can implement today.

Public Leaderboard

Compare your agent against others. See top performers by framework. Compete for the top spot.

Certification Badge

Embed a badge in your GitHub README showing your agent's score. Build trust with users.

Works Everywhere

Linux, macOS, Windows. LangChain, AutoGPT, CrewAI, Claude Code, or custom agents.

Works With Any Framework

🦜
LangChain
🤖
AutoGPT
👥
CrewAI
Claude Code
Custom Agents

Simple Pricing

Benchmark for free. Upgrade when you need continuous monitoring.

Benchmark

Free Forever

Test and score your AI agents

  • Unlimited benchmarks
  • Score 1-100 with letter grade
  • AI-powered recommendations
  • Shareable report link
  • Public leaderboard
  • GitHub README badge
  • Linux, macOS, Windows
  • Any AI framework
  • No account required
Install Free
For Teams

Dashboard

Free Beta

Continuous monitoring for teams

  • Everything in Benchmark
  • Real-time monitoring dashboard
  • Track all agents across platforms
  • Security alerts & threat detection
  • Cost tracking & budgets
  • Team collaboration
  • Historical analytics
  • Email/Slack notifications
  • Priority support
Try Dashboard

Get Started Now

Install in seconds. Run your first benchmark in under a minute.

Download for Windows macOS Linux

Or install via pip (all platforms):

Install via pip

# Install from PyPI (Linux, macOS, Windows)
pip install sudodog

Run Benchmark

# Start your AI agent first, then run:
sudodog-benchmark

# The tool will:
# 1. Detect running AI agents on your machine
# 2. Let you select which one to benchmark
# 3. Monitor it for 30 seconds
# 4. Give you a score and recommendations

Other Commands

# Scan for AI agents running on your machine
sudodog-scan

# Run an agent with continuous monitoring (connects to dashboard)
sudodog run python your_agent.py

# Integrate with Claude Code
sudodog integrate claude-code

Full Documentation • View on GitHub • MIT License

Need Continuous Monitoring?

The benchmark gives you a snapshot. The dashboard gives you real-time visibility across all your agents, all your platforms, with security alerts and cost tracking.

Try Dashboard Free