Type-safe queries for all 14 data models.
Auto-generated types from schema, multi-model joins, and migration management. One ORM for SQL, vectors, graphs, timeseries, documents, and more.
Core Features
Multi-Model Power
Search vector embeddings, join with SQL tables, and traverse graph relationships — all in a single fluent query. The ORM routes each part to the optimal storage engine automatically.
Insert a SQL row, store its vector embedding, and create graph relationships in one atomic transaction. If any step fails, everything rolls back. No partial states, no data inconsistency across models.
Language Support
Native implementations for every supported language, each designed to feel idiomatic.
Migration System
Version-controlled schema changes across all 14 data models. The ORM detects diffs between your schema and database, generates type-safe migration files (not raw SQL), and applies them with zero-downtime online schema changes.
Every migration includes a tested rollback function. Git-friendly files with conflict detection when multiple developers change schema simultaneously.
CLI Commands
nucleus migrate generate <name>Create migration from schema diffnucleus migrate upApply pending migrationsnucleus migrate downRollback last migrationnucleus migrate statusShow applied/pending migrationsnucleus migrate validateCheck migrations for errorsnucleus migrate resetRollback all migrations (dev only)Get Started
# Coming soon# Follow development at github.com/tystack