Search Pilla

Search pages and blog articles, or ask Pilla.

AI benchmarks

AI Benchmarks

We test the AI that powers Pilla on the real jobs it does for customers every day. Every published score comes from the same production model, with the cases and answers open for inspection.

Account admin
Managing users, teams, and account settings.
93.8%
avg. grade
Answers
Answering questions about the account in chat.
94.2%
avg. grade
Building workflows
Drafting a canvas workflow from a plain-English brief.
83.7%
avg. grade
Building automations
Wiring up automation steps on a canvas.
83.7%
avg. grade
Running workflows
Progressing a live workflow run to completion.
78.4%
avg. grade
Fetching data
Pulling account data to ground an answer or a decision.
88.9%
avg. grade
Making decisions
Yes/no judgment calls on a submitted piece of work.
97.9%
% correct
Taking actions
Composing a message to the team from workflow context.
95.7%
avg. grade