Learn everything here.
Run your first evaluation in 5 minutes.
Endpoint references and OpenRouter integration.
The High Council Protocol, performance-impact formula, and multi-evaluator consensus.
Agent profiles, certification, trust badges, and result logic.
Domain-expert agents interpret your documents — Medical, Legal, Education, Architecture & Construction.
Up to 4 users + 4 models in the same chat. Multiplayer mode for joint research, brainstorming, and social interaction.
Limits, models, and subscription details.
Measured output of the evaluation harness: method, datasets, results, limitations.