Jev vs Open Model Arena

YOUR LOCAL LEADERBOARD & COMPARATIVE ECONOMICS

Model Leaderboard vs TypeSafe Jev

Ranks models on decision closeness to Jev, speed, and cost efficiency in tokens and dollars. Results are saved only in this browser.

BENCHMARK BASELINE TypeSafe Jev (System One)

Token Pricing $0.042 / 1M input Output tokens free

Typical Cost / Run ~$0.000015 ~360 tokens

Typical Latency ~300 - 350ms Native calibration

Output Format Exact Probabilities Zero parsing errors

Rank	Model / Provider	Jev Fidelity (Closeness)	Speed / Latency	Tokens Used	Dollar Cost / Query	Cost vs Jev	Action

COMMON BENCHMARK INPUT

State & Evaluation Questions

The exact same state and questions are evaluated concurrently by both Jev and the chosen open model.

Presets:

State (Context to Evaluate) 0 chars

Evaluation Questions

Prompts are sent to TypeSafe and OpenRouter for evaluation. Benchmark results are saved in this browser only. There is no shared public leaderboard or account system.

SYSTEM ONE

TypeSafe Jev

Native Probabilistic Engine

Ready

Model Endpoint jev-latest (POST /v1/systemone)

Decision Type Calibrated probabilities, choices & scores

Prompt Engineering None required (native typed questions)

📊

Click Run Side-by-Side Comparison above to execute Jev against the open model.

{}

OPEN MODEL

Open Model

Emulating Jev via OpenRouter

Ready

Select Open Model (Curated Cheap & Fast):

🤖

Open model predictions will appear here side-by-side with Jev's output.

{}