Sold by: Isaacus
Deployed on AWS
State-of-the-art legal embedding and reranking models optimized for legal research, legal RAG, and legal document classification.
Overview
Kanon 2 Embedder and Kanon Universal Classifier are state-of-the-art legal embedding, reranking, and zero-shot classification models optimized for semantic search, RAG, and document classification. Kanon 2 Embedder can be used to sort through millions of legal documents to find the most similar passages to a user query, with Kanon Universal Classifier then reranking those passages by their relevance with extreme precision, making this model bundle especially valuable for legal research and legal RAG applications.
As of 28 October 2025, Kanon 2 Embedder ranks first on the Massive Legal Embedding Benchmark (MLEB) ahead of 20 other models, including OpenAI Text Embedding 3 Large, Gemini Embedding, Voyage 3 Large, Qwen 3 Embedding 8B, and Jina Embeddings v4 (https://arxiv.org/abs/2510.19365 ). It also ranks first on case, legislation, and regulation retrieval and third on contract retrieval.
Kanon Universal Classifier likewise manages to outperform its largest open-source equivalent, DeBERTa v3 large, at legal classification and NLI while remaining 71% faster.
Kanon 2 Embedder has a context window of up to 16,384 tokens, and Kanon Universal Classifier has a local context window of 512 tokens. Kanon Universal Classifier can process documents of any length thanks to Isaacus' semchunk semantic chunking algorithm (https://github.com/isaacus-dev/semchunk ).
On a single g6.2xlarge instance, Kanon 2 Embedder can embed up to ~15k legal documents (62 million tokens) per hour while Kanon Universal Classifier can process up to ~32k documents (131 million tokens) an hour.
Like all other Isaacus SageMaker models, your Kanon 2 Embedder and Kanon Universal Classifier bundle will be fully air-gapped--no data will enter or leave your AWS account.
Conveniently, Isaacus SageMaker models are also compatible with the standard Isaacus Python SDK via the Isaacus SageMaker Python integration (https://docs.isaacus.com/integrations/amazon-sagemaker ).
Both Kanon 2 Embedder (https://aws.amazon.com/marketplace/pp/prodview-qoz2rxyqhtewu ) and Kanon Universal Classifier (https://aws.amazon.com/marketplace/pp/prodview-6dotzeq7aq4sy ) can be purchased separately on the AWS Marketplace.
You can negotiate a discount to this model bundle by contacting us at https://isaacus.com/support .
Highlights
- Kanon 2 Embedder is ranked first on the [Massive Legal Embedding Benchmark (MLEB)](https://huggingface.co/papers/2510.19365) at legal document retrieval out of 20 other models, including OpenAI Text Embedding 3 Large, Gemini Embedding, Voyage 3 Large, Qwen 3 Embedding 8B, and Jina Embeddings v4, while Kanon Universal Classifier ranks ahead of its largest open-source competitors at legal natural language inference by both accuracy and inference time.
- Capable of embedding up to ~15k legal documents (62 million tokens) per hour and reranking up to ~32k documents (131 million tokens) per hour.
- Fully compatible with the Isaacus Python SDK via the [Isaacus SageMaker Python integration](https://docs.isaacus.com/integrations/amazon-sagemaker)--no substantive code changes necessary.
Details
Sold by
Delivery method
Latest version
Deployed on AWS
New
Introducing multi-product solutions
You can now purchase comprehensive solutions tailored to use cases and industries.
Features and programs
Financing for AWS Marketplace purchases
AWS Marketplace now accepts line of credit payments through the PNC Vendor Finance program. This program is available to select AWS customers in the US, excluding NV, NC, ND, TN, & VT.
Pricing
Pricing is based on actual usage, with charges varying according to how much you consume. Subscriptions have no end date and may be canceled any time.
Additional AWS infrastructure costs may apply. Use the AWS Pricing Calculator to estimate your infrastructure costs.
Dimension | Description | Cost/host/hour |
|---|---|---|
ml.g6.2xlarge Inference (Batch) Recommended | Model inference on the ml.g6.2xlarge instance type, batch mode | $7.99 |
ml.g6.2xlarge Inference (Real-Time) Recommended | Model inference on the ml.g6.2xlarge instance type, real-time mode | $7.99 |
ml.g5.2xlarge Inference (Batch) | Model inference on the ml.g5.2xlarge instance type, batch mode | $7.99 |
ml.g5.4xlarge Inference (Batch) | Model inference on the ml.g5.4xlarge instance type, batch mode | $7.99 |
ml.g5.8xlarge Inference (Batch) | Model inference on the ml.g5.8xlarge instance type, batch mode | $7.99 |
ml.g5.16xlarge Inference (Batch) | Model inference on the ml.g5.16xlarge instance type, batch mode | $7.99 |
ml.g6.4xlarge Inference (Batch) | Model inference on the ml.g6.4xlarge instance type, batch mode | $7.99 |
ml.g6.8xlarge Inference (Batch) | Model inference on the ml.g6.8xlarge instance type, batch mode | $7.99 |
ml.g6.16xlarge Inference (Batch) | Model inference on the ml.g6.16xlarge instance type, batch mode | $7.99 |
ml.g5.2xlarge Inference (Real-Time) | Model inference on the ml.g5.2xlarge instance type, real-time mode | $7.99 |
Dimensions summary
You pay by the hour for the SageMaker instance running the model, not by upfront commitment. Charges apply only while your endpoint is active. The 16 dimensions cover two instance families, g5 and g6, each in four sizes: 2xlarge, 4xlarge, 8xlarge, and 16xlarge. Larger sizes carry more compute for heavier workloads. Each size comes in two modes: Batch, for processing grouped inputs, and Real-Time, for live requests. You choose the family, size, and mode that fit your workload, and pay per host hour used.
Top-of-mind questions for buyers
One host hour is one hour that your SageMaker endpoint runs on the selected instance type. The instance provides the GPU and compute that hosts both bundled models. You are billed for each hour the endpoint stays active, regardless of how many requests you send during that hour.
Charges accrue per hour the endpoint stays active, not per request. An idle-but-running endpoint still bills host hours. To stop software charges, delete or shut down the endpoint. Underlying AWS resource fees may still apply until you remove the endpoint.
Batch mode processes grouped inputs together, suited to bulk jobs you run at intervals. Real-Time mode serves live requests as they arrive, suited to interactive use. Both bill per host hour on the same instance. You pick the mode matching your workload; each carries its own rate.
isaacus.com
Helpful?
Vendor refund policy
To the maximum extent permitted by law, there are no refunds for consumption of this product.
Legal
Vendor terms and conditions
Upon subscribing to this product, you must acknowledge and agree to the terms and conditions outlined in the vendor's End User License Agreement (EULA) .
Content disclaimer
Vendors are responsible for their product descriptions and other product content. AWS does not warrant that vendors' product descriptions or other product content are accurate, complete, reliable, current, or error-free.
Delivery details
Amazon SageMaker model
An Amazon SageMaker model package is a pre-trained machine learning model ready to use without additional training. Use the model package to create a model on Amazon SageMaker for real-time inference or batch processing. Amazon SageMaker is a fully managed platform for building, training, and deploying machine learning models at scale.
Deploy the model on Amazon SageMaker AI using the following options:
Version release notes
Patched release of this model with version 0.1.4 of the Isaacus SageMaker Model Server, fixing the inability to load models after 19 January 2026 due to an expired Isaacus license key that should now never expire.
Additional details
Inputs
- Summary
For a user-friendly walkthrough of how to get started deploying Isaacus models on SageMaker, check out the Isaacus SageMaker quickstart guide on our docs.
This model runs on the fully air-gapped Isaacus SageMaker Model Server, which supports all the same functionality as the standard Isaacus API except that requests to the server must be proxied through the /invocations endpoint.
For example, if you wanted to send a POST request to /v1/embeddings with the data {"model": "kanon-2-embedder", "texts": ["This is a confidentiality clause."], "task": "retrieval/query"}, you could so by sending /invocations the payload {"path": "/v1/embeddings","data": {"model": "kanon-2-embedder", "texts": ["This is a confidentiality clause."], "task": "retrieval/query"}}.
Likewise, if you wanted to send the request {"model":"kanon-universal-classifier","query":"Who is the Governor-General?","texts":["The Governor-General is Sam Mostyn.","The King is Charles III."]} to /v1/rerankings, you could do so by sending /invocations the payload {"path":"/v1/rerankings","data":{"model":"kanon-universal-classifier","query":"Who is the Governor-General?","texts":["The Governor-General is Sam Mostyn.","The King is Charles III."]}}.
This means that minimal code changes are necessary to switch between the online Isaacus API and your own private Isaacus model deployments.
In fact, Python users can use the Isaacus SageMaker Python integration to automatically forward requests to the Isaacus API to SageMaker deployments using the standard Isaacus SDK.
Given that this is a private deployment and that authentication is managed by AWS, Isaacus API keys are not needed and are ignored.
As an embedding model, Kanon 2 Embedder currently only supports the /v1/embeddings endpoint. As a reranking and classification model, Kanon Universal Classifier supports the /v1/rerankings and /v1/classifications/universal endpoints.
- Limitations for input type
- All the same limitations applicable to the Isaacus API except for the need for an API key.
- Input MIME type
- application/json
{
"path": "/v1/embeddings",
"data": {
"model": "kanon-2-embedder",
"texts": [
"Who was the plaintiff in Mabo?"
],
"task": "retrieval/query"
}
}
{"path": "/v1/embeddings","data": {"model": "kanon-2-embedder", "texts": ["This is a confidentiality clause."], "task": "retrieval/query"}}
{"path":"/v1/rerankings","data":{"model":"kanon-universal-classifier","query":"Who is the Governor-General?","texts":["The Governor-General is Sam Mostyn.","The King is Charles III."]}}
Input data descriptions
The following table describes supported input data fields for real-time inference and batch transform.
Field name | Description | Constraints | Required |
|---|---|---|---|
path | The path of the API endpoint being invoked (e.g., `/v1/embeddings`). | One of `v1/embeddings`, `/v1/rerankings`, `/v1/extractions/qa`, and `/v1/classifications/universal`. | Yes |
method | The HTTP method used for the invocation (e.g., `POST`). Defaults to `POST`. | One of `POST`. | No |
headers | The HTTP headers to include in the invocation request. Defaults to `null`/`None`, in which case no additional headers are sent. | Must be a mapping of strings to strings. | No |
data | The data to be sent as the body of the invocation request. This can be any serializable object. Defaults to `null`/`None`, in which case no body is sent. | - | No |
Resources
Vendor resources
Support
Vendor support
To get in touch with our support team, you can reach out via the support form on our website: https://isaacus.com/support . We endeavor to respond within 24 hours.
AWS infrastructure support
AWS Support is a one-on-one, fast-response support channel that is staffed 24x7x365 with experienced and technical support engineers. The service helps customers of all sizes and technical abilities to successfully utilize the products and features provided by Amazon Web Services.
Similar products
A state-of-the-art legal embedding model optimized for semantic search, RAG, and document classification, ranked first on the Massive Legal Embedding Benchmark.
Best-in-class legal enrichment, reranking, and embedding models, optimized for legal knowledge graphs and legal RAG.
A first-of-a-kind enrichment model, capable of transforming unstructured documents into rich, highly structured knowledge graphs.
A state-of-the-art reranking model optimized for legal RAG, research, and classification, ranked first on Legal RAG Bench and the Massive Legal Embedding Benchmark (MLEB).
Customer reviews
No customer reviews yet
Be the first to review this product . We've partnered with PeerSpot to gather customer feedback. You can share your experience by writing or recording a review, or scheduling a call with a PeerSpot analyst.