Misinformation spreads at internet speed. Truth takes work, sources, and careful reasoning.
Lenz exists to close that gap — not truth as we see it, but truth as the evidence supports it.
A process, not an opinion
Lenz doesn’t ask one AI model what it thinks and call it a day. Every claim runs through a structured pipeline: framing, independent research, a multi-model debate, a panel of adjudicators, and finally a human review. Each step is designed to catch what the previous one missed.
Lenz is powered by AI, and AI makes mistakes — confidently, fluently, and without hesitation. We know that. But a structured process with multiple models cross-examining each other, grounded in independently retrieved sources, is meaningfully harder to fool than a single model guessing from memory.
Not perfection, but a better process. See how it works →
What we believe
- Evidence first, opinions never. Conclusions follow the sources, not the other way around.
- Transparency builds trust. You can see the research, the debate, the scores — everything.
- Multiple perspectives beat a single voice. Different models, different training data, different blind spots.
- Humans stay in the loop. Every claim published in the Library is reviewed by a human editor.
Who this is for
Students writing papers. Parents checking health advice. Journalists on deadline. Anyone who’s ever paused before hitting share.
If you’ve ever thought “is this actually true?” — Lenz is built for that moment.
Who we are
A small team of human enthusiasts and a few AI agent coworkers. Self-funded and independent — we built Lenz because we wanted it to exist.