Lenz IO
Lenz is an independent verification layer and fact-checking API for AI-generated and human-written documents, answers, and reports. It checks factual claims against independent sources through a multi-stage, multi-model adversarial pipeline and returns a scored, cited verdict with a full audit trail.
At a Glance
- AI product teams
- Developers and teams shipping AI-generated content
- Publishers and communications teams
- Research and OSINT teams
- +3 more
AI Tools by Lenz IO
(1)
Lenz
AI Hallucination Detection API
Discussions
No discussions yet
Be the first to start a discussion about Lenz IO
Latest News
Lenz's second Product Hunt launch as an independent, multi-model fact-checking API; ranked #3 of the day.
Lenz Research published its study of frontier-model disagreement on 1,000 recent claims, reporting 63% panel disagreement and 23% substantial disagreement among usable cases.
Published 'Three claims, three verdicts, one process,' an example of statement verification across domains.
First Product Hunt launch: 'Fact-check any statement with source-backed, multi-model AI.'
Products & Services
REST API for extracting claims, quickly assessing them, running a full sourced verification, and asking follow-up questions grounded in a completed verification. The deep pipeline covers framing, research, adversarial debate, panel review, and conclusion.
A public workbench where a user can paste a statement, link, or question and receive a sourced verdict with the reasoning shown.
Free endpoint that pulls atomic, verifiable claims from text; the developer documentation lists a quota of 1,000 calls per day on the free tier.
Fast three-model panel assessment, returning claim-level verdicts and confidence in roughly 5–10 seconds.
Market Position
Lenz positions itself as an independent verification layer for AI products, differentiated from simply asking one LLM for an answer or relying only on RAG groundedness checks. Its distinction is the combination of independent web evidence, multi-vendor models, explicit for/against debate, independent review axes, scored verdicts, and an auditable citation trail. Adjacent alternatives include single-model AI fact-checking and generic LLM/RAG evaluation workflows.
Leadership
Founders
Kosta Jordanov
Founder & CEO of Lenz; public LinkedIn search results identify him as based in Bulgaria, educated in computer science at the Technical University of Sofia, and formerly a founder at Saltanat Labs.
Executive Team
Kosta Jordanov
Founder & CEO
Founder and CEO of Lenz; public profile results identify prior founder experience at Saltanat Labs and computer-science study at the Technical University of Sofia.
Founding Story
Kosta Jordanov says Lenz was started because businesses increasingly ship AI-generated content and need a way to verify factual claims before that content reaches customers. The initial vision was to package source-backed, multi-model verification as an API/SDK and make it accessible across common platforms, rather than relying on a single model's memory or answer.
Business Model
Revenue Model
Self-serve subscriptions with usage quotas for API and integration calls; enterprise pricing is customized for volume beyond Scale, SLAs, white-labeling, and custom integration support.
Pricing Tiers
1,000 extractions/day, 100 fast checks/month, 10 deep verifications/month, and 20 follow-ups/month.
For individuals; 500 fast checks/month, 50 deep verifications/month, and 100 follow-ups/month.
For developers shipping verified AI content; 5,000 fast checks/month, 500 deep verifications/month, and 1,000 follow-ups/month; $999/year annual option.
For production integrations; 20,000 fast checks/month, 2,000 deep verifications/month, and 4,000 follow-ups/month; $3,990/year annual option.
Volume beyond Scale, SLAs, white-label, and custom integration support.
Target Markets
- AI product teams
- Developers and teams shipping AI-generated content
- Publishers and communications teams
- Research and OSINT teams
- Regulated industries and compliance-sensitive organizations
- Customer-support and agent-platform teams
- Runtime gating of outbound AI text before it reaches users
- Pre-release and CI regression checks for AI-generated content
- Incident triage when a customer flags a potentially incorrect answer
- AI-written newsletters, blog posts, product updates, and other publishing workflows
- Data enrichment and CRM claim verification
- Regulated health, finance, supplement, and other high-stakes claims