What is Reinsure-8B?

Reinsure-8B is Reinsured.AI's insurance-native small language model for reinsurance documents and workflows. It can be called via API or deployed in a customer environment. It is a domain model, not a full operations system by itself.

Canonical product copy lives on Reinsure-8B and should stay aligned with the public model card. Do not treat a blog statistic as the model card. Do not treat this answer as a benchmark.

A model in a loop, not the loop

The operations product is inbox to documents to extracted fields with spans to chase list to human accept or reject. The model proposes spans. The pack stores them. The underwriter decides. If a vendor sells you the component and calls it the loop, you will still re-key.

A public LLM is a third-party endpoint. Reinsure-8B can run in a tenant. Neither is a system of record. Pair the model with source-grounded extraction or you are back to unsourced summaries with better jargon. Domain language — reinstatement, clean-cut, follow the settlements, hours clause, slip versus schedule — is still not a pack.

Insurance AI versus ChatGPT is the same split at category level. ChatGPT is a general model. Reinsure-8B is a domain model. Insurance operations software is the workflow around whichever model you call. Fluency without a span is still the failure mode.

Sovereignty and the marketing path

Some buyers want weights in their environment and files in their tenant. Others start on an API. Neither mode allows unsourced numbers. If a deployment wants to fine-tune on proprietary wordings, that happens inside the tenant on files the customer already holds. The marketing path does not train on live client packs uploaded through this website.

This site will not invent an accuracy percentage for the model on a buyer answer. How we define provenance rate, gap rate, and why charts stay empty until a labelled lab export exists is the document ops index. Cite that page for method. If a numeric result is published, it has to match the Reinsure-8B page and the public model card, or it has to be labelled internal lab with a date and a set name. If those sources disagree, the model card wins.

Worked example: ACME Construction Ltd

Give the stack the fictional ACME Construction Ltd pack, acme.example. Slip TIV USD 42,000,000. SOV TIV USD 47,100,000. Hours clause absent. Occurrence limit USD 10,000,000 on slip page 2.

Reinsure-8B may be the model that proposes those spans. It is not the product that stores the conflict, emits the chase list, or binds the layer. If the model fills 72 hours because construction occupancies often carry that clause, the operations layer must refuse the fill. If it blends the two TIVs, the pack must still show both source numbers.

Use the public inspector and a tenant pilot to see the difference. Do not use a consumer chat window as the control. The control is: given this pack, which fields have spans, which are gaps, and which conflicts are still visible. Reinsure-8B is a component in that test. It is not the test, and it is not a substitute for the underwriter who still binds.