Editor pick · Visual and animation
Pelican test
A repeatable visual and animation prompt that makes structure, contact, and motion easy to inspect side by side.
View evaluationEvaluations / Test directory
Tests you can inspect, compare, and run
Explore focused evaluation suites with their method, evidence, related writing, and a clear path to run a test yourself.
01 / FEATURED EVALUATIONS
Start with a question, not a score.
Each suite has a defined scope. Open its detail page to see the method, evidence boundary, and the related notes before you draw a conclusion.
Editor pick · API diagnostics
API quality check
A private, one-time check for protocol shape, response integrity, model claims, and bounded timing signals.
View evaluation
Editor pick · Station operations
Station reliability
A longer-running method for connections, timeouts, latency, retries, and changes across a named observation window.
View evaluation02 / ALL EVALUATIONS
Choose the test that fits your question.
The directory separates visual experiments, private diagnostics, and longer-running station checks so their results are not treated as the same kind of evidence.
Visual and animation
Experiment / galleryPelican test
A repeatable visual and animation prompt that makes structure, contact, and motion easy to inspect side by side.
- Status
- Experiment / gallery
- Results
- Gallery available
API diagnostics
Run it yourselfAPI quality check
A private, one-time check for protocol shape, response integrity, model claims, and bounded timing signals.
- Status
- Run it yourself
- Results
- Results not published yet