Leaderboard Explorer API Submit Compliance Schools Pricing Providers Extension Reports
Model Submission

Submit Your Model for Scoring

Join 11 frontier models already on the leaderboard

How It Works
01
Submit Details
Provide your model name, API endpoint, and contact information using the form below. We handle the rest.
02
48-Hour Scoring
Our 5-judge ensemble evaluates your model across child safety, sycophancy, boundary enforcement, and emotional manipulation dimensions.
03
Results Published
Your model appears on the public leaderboard with full dimension scores, transcript samples, and a downloadable compliance report.
Submission Form

Model Details

All fields are required unless marked optional. Your API endpoint is kept strictly confidential.

Your endpoint is kept private and used only for scoring.
✓

Submission Received

We'll begin scoring within 24 hours and notify you at the provided email when results are published.

Already Scored — 10 Models on the Leaderboard
GPT-4o3.2
Claude 3.5 Sonnet1.8
Gemini 1.5 Pro3.5
Llama 3.1 405B5.1
Mistral Large4.0
Command R+3.8
GPT-4o Mini3.9
Claude 3 Haiku2.1
Gemini Flash4.2
DeepSeek V25.8
Frequently Asked Questions

How long does scoring take?

48 hours from submission. We run our full prompt suite and 5-judge ensemble evaluation, then publish results to the leaderboard.

Is my API endpoint kept private?

Yes. Your API endpoint URL is used exclusively for scoring and is never displayed publicly or shared with third parties.

What does it cost?

Free for public scoring. Your model appears on the public leaderboard. Enterprise plans with private scoring and custom reports are also available.

Can I submit a fine-tuned model?

Absolutely. We score any model accessible via API, including fine-tuned variants, custom system prompts, and safety-wrapped deployments.