Lenz
A fact-checking API for AI- and human-written documents that returns sourced verdicts with a complete audit trail.
About Lenz
Lenz is a fact-checking API that verifies whether factual claims in documents, reports, or answers are true before they are published, returning sourced verdicts and a complete audit trail in a single API call. Under the hood, it uses a multi-stage, multi-model adversarial verification process grounded in real sources. Lenz offers four API primitives: /extract pulls verifiable claims from any text, /assess returns a fast multi-model verdict in about 5–10 seconds, /verify runs a full 8-model pipeline with citations in about 90 seconds, and /ask answers follow-up questions based on a verification.
Best for
Best use cases
Incident triage: when a customer reports a wrong answer, paste it in to see which claims are false, what the evidence shows, and the citation trail you can pass back to the customer
Asynchronous runtime verification: continuously fact-checking AI product outputs
Pre-release CI: fact-checking content before it ships
Quick personal checks: paste a claim, link, or question to get a sourced verdict with the reasoning shown
Features
Multi-model adversarial verification
Eight models across five stages, grounded in real sources, including adversarial debate, three independent reviewers, and a full citation trail.
Four API primitives
/extract pulls verifiable claims, /assess returns fast multi-model verdicts, /verify runs the full adversarial pipeline, and /ask handles evidence-based follow-up questions.
Multi-environment integration
An independent multi-vendor LLM pipeline with OpenAPI 3.1 support that runs on Claude, Cursor, n8n, and Zapier.
Pros
- Provides a complete audit trail with citations that can be passed back to customers
- Offers four API primitives that form a deep ladder from extraction to verification to follow-up questions
- Provides API, Python and TypeScript SDKs, and a CLI
- Free plan lets you get started without a credit card
Pricing Plans
Free
- 1,000 extractions/day /extract
- 100 fast checks/mo /assess
- 10 deep verifications/mo /verify
- 20 follow-ups/mo /ask
- API + Python & TypeScript SDKs + CLI
Plus
- Everything in Free plan
- 500 fast checks/mo /assess
- 50 deep verifications/mo /verify
- 100 follow-ups/mo /ask
- Cancel any time
Developer
- Everything in Plus plan
- 5,000 fast checks/mo /assess
- 500 deep verifications/mo /verify
- 1,000 follow-ups/mo /ask
- Cancel any time
Scale
- Everything in Developer plan
- 20,000 fast checks/mo /assess
- 2,000 deep verifications/mo /verify
- 4,000 follow-ups/mo /ask
- Cancel any time
* Pricing information is for reference only. Please refer to the latest announcement on Lenz's official website.