Fact API for language models
Veridical Intelligence is a natural-language API that your LLM can ask about facts. Every answer comes back with references, so your model can cite what it says instead of making it up. Our goal is to eliminate hallucination.
Get API accessTap a highlighted phrase in this example answer to see the reference behind it, or the lack of one.
Q: What is our refund window for annual plans?
Annual plans can be refunded within 30 days of purchase, prorated for unused months, and refunds are processed within 24 hours.For teams putting language models in front of customers, analysts, or decision makers.
A natural-language API your model can talk to about facts. It answers in plain language and returns the references behind each answer.
We work with your engineers to connect the API to your models and applications, so grounded answers reach your users quickly.
The API sits between your model and the answer your users see.
When your LLM needs a fact, it asks the Veridical API in plain language.
The API replies with the fact and the references that support it.
Your model builds its response from referenced facts, not from guesses.
Every claim your users read can be traced back to its source.
Our team brings decades of software engineering experience and years of hands-on work with AI. We built Veridical Intelligence because we think language models are only as useful as they are truthful.
Tell us what you're building, and we'll help you get started with the API.