Isaacus

Hallucination-free legal AI models. Really.

The most accurate and scalable frontier legal AI models. Own your data; as you should.

Get building Book a demo

We lead the pack in frontier legal AI.

Find the right passages in contracts, briefs, and dockets.

Kanon 2 Embedder powers semantic search; Kanon 2 Reranker scores and sorts documents by relevance to your query.

Our products are used by

  • Microsoft
  • World Bank
  • IBM
  • University of California, Berkeley
  • The University of Sydney
  • The University of Melbourne
  • Monash University

We meet you where you are.

Nobody enjoys being locked into a single deployment mode, so why do that to your clients?

Online platform

Access our models and documented APIs via our online platform, with continuous availability, minimal latency, and zero permanent data retention. See the documentation or quickstart guide.

Private deployments

Host our models and APIs in your own AWS tenancy or on your own hardware with private, air-gapped containers—ensuring maximal SOC, ISO, and IRAP compliance. Learn about Amazon SageMaker deployments.

Trusted by legal teams

From research platforms to government agencies and AI-native startups.

“After extensive testing across multiple legal queries, Isaacus consistently outperforms the fine-tuned Cohere model for legal document reranking. Isaacus demonstrates a deeper understanding of legal concepts, hierarchies, and relationships between different types of legislation.”
Hamish Cameron, Legal Oracle
“We found the speed and accuracy of Kanon 2 Enricher to be amazing!”
Spinal, Accounts receivable AI
“Its confidence scoring provides clearer signals about relevance, and it shows superior performance for specialised legal domains.”
Hamish Cameron, Legal Oracle
“The introduction of Isaacus legal AI models to the AWS Marketplace will allow us to rapidly and securely deploy Isaacus models across multiple Australian government departments.”
Hamish Cameron, AccuFind Law
“The most compelling evidence is how Isaacus leverages enhanced context to produce rankings that better align with legal research best practices.”
Hamish Cameron, Legal Oracle

From the blog

The latest on our models, research, and benchmarks.