On every plan
Agents read the governed model through the MCP endpoint on Builder, Team and Enterprise. Agents are not seats; viewers are free and unlimited.
What a Core Model can do once it exists
Graph-based modeling, adjustable meta-models, collaboration and suggestions, import-map-compare, precision editing, validation and governance, transformation across formats, and the agent endpoint.
An inventory, without adjectives. Everything here is in the product today; where a capability is early access, the connector page says so.
What a Core Model can do
Six capabilities of the modeling workspace, as they look in the product.
Graph-Based Low-Code Modeling Platform
CoreModels provides visual editors for entities, relationships, and constraints — letting you build and evolve data models without writing code. Everything is connected in a single schema graph.
Adjustable Meta-Models
CoreModels comes with default meta-model templates for common formats — all customizable. Extend existing templates or build your own to match your organization's needs.
- •JSON Schema for API and service definitions
- •JSON-LD and linked-data models for rich semantics
- •Configurations for transformations and orchestration
- •Custom formats for organizational needs
Effortless Collaboration & Suggestions
Collaborate with your team on data models in real time. Get feedback, make changes, and ensure everyone is on the same page with built-in suggestions and comments.
Import, Map & Compare
Seamless Ingestion: Bring in models from existing systems or upload new definitions via simple import flows.
Mapping & Comparison: Align disparate schemas, highlight differences, and reconcile mismatches quickly.
Expandable Export: Share refined models in the formats you need — documentation, code generation, or integration.
Precision Editing & Customization
Fine-tune your data models to meet specific needs. Specify required attributes, create custom relationships, and control every detail of your schema definitions.
Validation & Governance
Rule Definition: Specify required attributes, custom relationships, and constraints.
Automated Validation: Catch errors before they propagate, guaranteeing schema integrity across pipelines.
Audit Trails: Maintain transparent records of who changed what and when, ensuring compliance and trust.
Built as a cloud-native service, CoreModels scales to handle large datasets and complex model libraries. It integrates smoothly into your existing stack and keeps performance high as your usage grows.
Transformation across formats
Underneath every import and export is a transformation engine with one neutral representation at its center. A model that arrives as a JSON Schema can leave as a JSON-LD context, a ShEx shape, an Avro record, a SQL DDL, or an OWL class — and the engine keeps a ledger of what each target format cannot hold, so a lossy projection is declared rather than discovered. The same engine reconciles a schema imported from one platform against the same schema imported from another.
The agent endpoint
The person’s permissions
An agent runs under the credentials of the person who connected it and can do what they can do — read, and write within the same limits.
Every write is a version
An agent’s changes are versioned like anyone else’s, so they are reviewable and reversible. Agent prompts and traffic are not used to train models.
Bounded traffic
Each agent gets up to 15 concurrent in-flight requests; above that it receives a rate-limit response and retries. Enterprise plans set throughput in contract.
A library of connectors
16 platforms and 10 formats today, labelled for what you can do with each. Platforms are the systems your schemas run in; formats are the languages they are written in. Every card on the catalog says whether it is generally available, early access, or guided.
Platforms
Airbyte · Apache Airflow · Google BigQuery · cBioPortal · Confluent · Schema Converter · Databricks · dbt · Microsoft Fabric · AWS Glue · Neo4j · Open Semantic Interchange · REDCap · Salesforce · Snowflake · Azure Synapse
Formats
Apache Avro · JSON-LD · JSON Schema · LinkML · ODCS · MACH ODM · OWL · Protocol Buffers · ShEx · SQL
Built as a service
Built as a cloud-native service, CoreModels scales to large datasets and complex model libraries and keeps performance high as usage grows. Builder allows 100,000 nodes per model, Team 1,000,000, Enterprise unlimited — a node being any type, element, taxonomy term or value in a model; a 300-model dbt project at about eight columns each is roughly 2,400 nodes before vocabularies. Enterprise adds RBAC, SSO/SAML, IP allowlisting, audit logs, private schema vaults, dedicated infrastructure, on-premise and private-cloud deployment, and a 99.9% uptime commitment in contract. Plans and limits →
Try it
Next in the frameworkWho it serves →Six kinds of team keep meaning somewhere. One model lets them keep it together — data and pipeline, standards and ontology, API and interface, content and semantics, research and domain, business systems and contracts.Start your 14-day trial
Nothing is charged during the trial. Plans and limits are on the pricing page.
Previous: How it works