Ask anything. Auto-Route sends your question to the 3 AIs strongest on that topic — and shows you all three, side by side.
One AI can be confidently wrong. Three AIs answering the same question turns disagreement into a signal — the moment two diverge, you know something's worth checking.
"When 2 of 3 disagree, you've just caught something a single chatbot would have hidden."
The comparison is the product. Auto-Route doesn't collapse three answers into one — it picks which three to compare, then shows you all of them.
Auto-Route reads your question, identifies the topic, and selects the 3 engines that perform best on it from the routing matrix. It's the difference between "any 3 AIs" and "the right 3."
Auto-Route picks which engines to compare — it never collapses answers to one. That single-answer job belongs to the Quick-Claim Verdict at the bottom of your results.
And you stay in control: toggle Auto-Route off any time to choose engines yourself.
Auto-Route is included with the 7-day free trial and the Smart plan ($9/mo).
The routing matrix is a live scoreboard of which AIs perform best at which topics — coding, medical, legal, finance, science, current events, and more. It's the data Auto-Route reads to make its pick.
We run real AI-vs-AI comparisons continuously in the background, and the matrix updates as the evidence accumulates. No hand-curation, no pay-to-rank.
Free accounts see an animated preview of how the matrix works. The full live matrix — with real, continuously-updating rankings — unlocks at Smart ($9/mo).
AI models update. Evidence accumulates. A model that led on medical questions last month may have been overtaken by a competitor that just shipped a new version.
"Yesterday's screenshot is yesterday's truth — the matrix you rely on is always the current one."
The value isn't a cheat-sheet you memorize. It's always-current intelligence. A ranking that shifts did so because the underlying data shifted — not because we changed anything. That's what keeps it honest.
Numbers that flicker sub-second are a gimmick. Numbers that drift day-to-day because real comparisons came in — that's the truth signal. We run background comparisons specifically so the matrix reflects evidence, not a static snapshot.
| Tier | Matrix access | Engines |
|---|---|---|
| Free | Animated preview — how it works | ChatGPT · Gemini · Meta Llama (free trio) |
| 7-day trial | Full live matrix + Auto-Route | All Smart AIs (no CC required) |
| Smart · $9/mo | Full live matrix + Auto-Route | + Claude Haiku · Sonnet · GPT-4o · Gemini Pro |
| Super Smart · $19/mo | Full live matrix + Auto-Route | + Claude Opus · OpenAI o3 · Wildcards ⚠ |
Does Auto-Route hide engines from me?
No. You always see all 3 results side by side, and you can toggle Auto-Route off any time to choose the engines yourself.
How are rankings decided?
Real, continuously-run AI-vs-AI comparisons — no hand-curation, no sponsored rankings. If a model falls behind, the score reflects it.
Why did a ranking change since last week?
Because it's live. AI models update frequently, and as new evidence comes in from our background comparison runs, rankings shift to match.
Is my question kept private?
Engine calls are single-turn with no stored conversation history — each search is a fresh, independent call. See our Privacy Policy for full details.