The trust layer for the AI universe
One benchmark.
Every AI platform.
A single, independent score that rates AI platforms on real, comparable criteria, updated continuously. Built for the age of infinite tools.
Scorecard
The Scorecard
Get the full pilot scorecard
All 20 platforms, every category, delivered to your inbox. Plus monthly benchmark notes. No spam, ever.
Methodology
How the score is built
Each platform is evaluated across seven weighted categories, each scored 0–100, then combined into a single weighted composite. Drag the sliders to see how weight changes affect scores.
Read the full methodology →Composite Formula
Steiner Index = Σ (category score × category weight)
Capability
Output quality, task accuracy, and benchmark performance.
Reliability
Uptime, latency consistency, and incident transparency.
Privacy
Data handling, opt-outs, certifications, and reporting.
Pricing
Clarity of pricing, hidden costs, and value for the tier.
Ecosystem
API quality, integrations, and developer tooling.
Update Velocity
Frequency and substance of meaningful updates.
Community
Documentation, support responsiveness, and community health.
Total Weight
100%
Governance: No platform can pay to influence its score. Any commercial relationship between Steiner Index and a scored platform must be publicly disclosed on that platform's score page.
Blog
Field Notes from the AI Benchmarking Frontier
Why AI Needs a Moody's
Steiner Index Team · August 2026 · 4 min read
AI has no Moody's, no Nielsen, no shared yardstick. That structural gap is the problem Steiner Index exists to solve.
Read more
How We Score: The 7-Category Methodology
Steiner Index Team · August 2026 · 6 min read
Seven weighted categories, one transparent composite. Every weight public, every data point traceable.
Read more
The $770K Pre-Seed: What We're Building in 4 Weeks
Steiner Index Team · September 2026 · 5 min read
$770K and four weeks to a live, investor-ready demo. Here is the full plan, budget line by budget line.
Read moreJoin the Steiner Index list
Score updates, new platforms, and methodology notes, once a month. Unsubscribe anytime.