llms.txt
Compare with other sites' llms.txt files →# Isaacus
> Hallucination-free legal AI models and APIs for developers — embedding, reranking, extraction, classification, enrichment, and knowledge graphs.
Isaacus builds frontier legal AI infrastructure. Use the links below for human-readable pages, or fetch the full docs corpus in one file for AI tooling.
## Documentation
- [Full documentation (isaacus.com/llms-full.md)](https://isaacus.com/llms-full.md): Every docs page concatenated as Markdown
- [Docs home](https://isaacus.com/docs): Developer documentation for Isaacus legal AI APIs
- [Quickstart](https://isaacus.com/docs/quickstart): Get started with Isaacus in minutes
- [API reference](https://isaacus.com/docs/api-reference/making-requests): Complete API reference for Isaacus legal AI models
## Product
- [Home](https://isaacus.com/): Isaacus is legal AI research company building legal superintelligence. Try out our hallucination-free models.
- [Kanon models](https://isaacus.com/products/kanon): Kanon is a family of legal AI models for embedding, reranking, extraction, enrichment, and classification
- [Embedding](https://isaacus.com/products/embedding): Generate state-of-the-art legal embeddings with Kanon 2 Embedder for semantic search, legal RAG, clustering, and retrieval across documents.
- [Reranking](https://isaacus.com/products/reranking): Improve legal search accuracy with Kanon 2 Reranker, ranking the most relevant legal passages first for RAG, document search, and AI assistants.
- [Extraction](https://isaacus.com/products/extraction): Extract precise answers from legal documents with a hallucination-free legal extraction model designed for contracts, legislation, judgments, and evidence.
- [Classification](https://isaacus.com/products/classification): Automatically classify legal documents and clauses using Kanon Universal Classifier with zero-shot legal document classification and no model fine-tuning.
- [Enrichment](https://isaacus.com/products/enrichment): Extract entities and relationships from legal documents to build searchable legal knowledge graphs for AI agents, document intelligence, and legal research.
- [Pricing](https://isaacus.com/pricing): View Isaacus pricing for legal AI models including embedding, reranking, extraction, classification, and research APIs. Start free or deploy privately.
## Docs pages
- [Welcome to the Isaacus Docs](https://isaacus.com/docs): Developer documentation for Isaacus legal AI models, APIs, SDKs, integrations, and guides for building accurate, citation-grounded legal AI applications.
- [Quickstart](https://isaacus.com/docs/quickstart): Get started with Isaacus in minutes. Make your first API request and build legal AI using our embedding, reranking, extraction, and classification models.
- [Capabilities](https://isaacus.com/docs/capabilities/introduction): Learn about the core capabilities of Isaacus, including legal embedding, reranking, extraction, enrichment, and zero-shot classification models.
- [Enrichment](https://isaacus.com/docs/capabilities/enrichment): Extract legal entities and relationships to build structured knowledge graphs using Isaacus enrichment models.
- [Embedding](https://isaacus.com/docs/capabilities/embedding): Generate legal embeddings for semantic search, retrieval, clustering, and RAG using Isaacus legal embedding models.
- [Reranking](https://isaacus.com/docs/capabilities/reranking): Improve legal retrieval quality with Isaacus reranking models that rank the most relevant passages for legal search and RAG.
- [Extractive question answering](https://isaacus.com/docs/capabilities/extractive-question-answering): Extract precise answers directly from legal documents with grounded extractive question answering models that avoid generative hallucinations.
- [Universal classification](https://isaacus.com/docs/capabilities/universal-classification): Classify legal documents, clauses, and text with zero-shot universal classification models requiring no task-specific training.
- [Our models](https://isaacus.com/docs/models/introduction): Learn how Isaacus legal AI models work, including architecture, capabilities, supported tasks, and deployment options.
- [Tokenization](https://isaacus.com/docs/models/tokenization): Understand how Isaacus tokenizes legal documents and prepares text for embedding, retrieval, reranking, and downstream AI tasks.
- [Chunking](https://isaacus.com/docs/models/chunking): Learn semantic chunking best practices for legal AI using semchunk to preserve legal context for retrieval and RAG applications.
- [Isaacus on Amazon SageMaker](https://isaacus.com/docs/integrations/amazon-sagemaker): Deploy Isaacus legal AI models on Amazon SageMaker for scalable, secure, and enterprise-ready inference.
- [Isaacus on LangChain](https://isaacus.com/docs/integrations/langchain): Use Isaacus legal embeddings with LangChain in Python via the langchain-isaacus integration package.
- [Isaacus Legal Document Graph Schema (ILDGS)](https://isaacus.com/docs/ildgs/introduction): Learn about the Isaacus Legal Document Graph Schema (ILDGS), the open schema used to represent structured legal knowledge.
- [Isaacus Query Language (IQL)](https://isaacus.com/docs/iql/introduction): Introduction to Isaacus Query Language (IQL), a language for querying legal knowledge graphs and structured legal information.
- [IQL templates](https://isaacus.com/docs/iql/templates): Explore reusable Isaacus Query Language templates for common legal search, retrieval, and graph queries.
- [IQL specification](https://isaacus.com/docs/iql/specification): Complete reference for the Isaacus Query Language specification, including syntax, operators, grammar, and supported query patterns.
- [Making requests](https://isaacus.com/docs/api-reference/making-requests): Learn how to authenticate, send requests, and work with Isaacus APIs using HTTP, SDKs, and production best practices.
- [Errors](https://isaacus.com/docs/api-reference/errors): Reference every Isaacus API error code with explanations, troubleshooting guidance, and recommended handling strategies.
- [Universal classification](https://isaacus.com/docs/api-reference/classifications/universal-classification): API reference for the Isaacus Universal Classification endpoint, including parameters, responses, and code examples.
- [Embedding](https://isaacus.com/docs/api-reference/embeddings/embedding): API reference for generating legal embeddings with Isaacus, including request parameters, responses, and example integrations.
- [Enrichment](https://isaacus.com/docs/api-reference/enrichments/enrichment): API reference for extracting structured legal entities and relationships using the Isaacus Enrichment API.
- [Extractive question answering](https://isaacus.com/docs/api-reference/extractions/extractive-question-answering): API reference for grounded extractive question answering over legal documents with the Isaacus Extraction API.
- [Reranking](https://isaacus.com/docs/api-reference/rerankings/reranking): API reference for reranking legal search results with the Isaacus Reranking API, including parameters and example requests.
## Optional
- [Blog](https://isaacus.com/blog): Engineering articles on legal AI, retrieval, RAG, semantic search, benchmarks, product releases, and building reliable AI systems for legal professionals.
- [MLEB benchmark](https://isaacus.com/mleb): Compare legal embedding models using the Massive Legal Embedding Benchmark, evaluating retrieval quality across legal documents.
- [About](https://isaacus.com/about): Isaacus builds trustworthy frontier legal AI infrastructure including embedding, reranking, extraction, classification, and knowledge graph models.
- [Support](https://isaacus.com/support): Get help with the Isaacus platform, APIs, billing, documentation, and legal AI models through our support resources and developer guides.