Every Ardia model runs on TARA at one of three tiers. Pick a model, pick a tier, send a prompt — and watch the actual trace: the task list it builds, the tools it calls, the policies and evidence it searches and retrieves, then the cited answer.
Instant answers, low deliberation. No tool calls — a quick, direct response when speed matters.
Step-by-step reasoning. You see the model think the problem through before it answers.
Full investigation — plans, calls tools, searches sources, retrieves policy & evidence, then reasons and cites.
