New·See the latest water task rankings
Water Benchmarks

Get Your Model on the Neutral, Expert-Led Benchmark

Evaluated by practitioners, not analysts. Submissions never influence scoring.

Submit

Get your model benchmarked

Tell us about your product and what you want evaluated. Low-friction intake first — no API secrets here.

Access for testing *
What outcome are you looking for? *

By submitting, you agree to our privacy policy.

How it works

1

Submit

Send model metadata and a contact through the form above.

2

Validate

We confirm access and configuration (≈2 days). API credentials are shared encrypted with PGP in a later step.

3

Evaluate

Your model runs the full case set and is scored by the validated judge panel (≈10 days).

4

Publish

Your public leaderboard row goes live — pass-rate, calibration, cost and status.

What you get

Public leaderboard row

  • •Pass / Conditional / Fail
  • •Q, Brier, ECE
  • •Cost & latency
  • •Eligible / Disqualified status
Get benchmarked

Private report

  • •Per-case breakdown
  • •Failure analysis
  • •Calibration of your model
  • •Standard & Premium tiers
Get benchmarked

Water Benchmarks Certified

  • •Signed certification seal
  • •For utilities & buyers
  • •Verifiable eligibility
  • •Requires a passing run
Get benchmarked
Neutrality is the asset. Submissions never influence scoring. The leaderboard and methodology stay fully open; commercial services live alongside the ranking, never on top of it.