Feature status · Python

The Python SDK is in beta. This table tracks coverage against the TypeScript packages.

Feature	TypeScript	Python
Dimensions and base measures	Available	Available
Filtered measures	Available	Available
Filters on dimensions and declared filters with operator allowlists	Available	Available
Null-check filters (`isNull`, `isNotNull`)	Available	Planned. The protocol accepts them; Python datasets can't write them yet
Relationships (`belongs_to`, `has_one`, `has_many`)	Available	Available
Time grains	Available	Available, including per-dataset `time_grains` restriction
Sync and async execution	Available	Available
Result caching and multi-tenancy	Available	Available
Dataset catalog	Available	Available
Typed DataFrames (`to_pandas`, `to_polars`, `to_arrow`)	Not applicable	Available
Framework integrations	Next.js, Node.js, Vite guides	Django, Flask, SQLAlchemy (`hypequery.contrib`)
`approxCountDistinct` measures	Available	Planned
`measure.merge` over `AggregatingMergeTree` state columns	Available	Planned
`final: true` for `ReplacingMergeTree` / `CollapsingMergeTree` sources	Available	Planned
Derived measures	Available	Available (`measure.derived` with `formula.*`)
Window measures (trailing, to-date, cumulative)	Available	Planned
Period comparisons (`shift`)	Available	Planned
Segments	Available	Planned
Relationship checks	Available	Available (`check_relationships`)
Cloud publishing	Not yet	hypequery cloud is not live. Bundle export works today (`write_dataset_bundle`); one-step publish is planned
MCP server for AI agents	Available	Through the Node `hypequery mcp` bridge over the server's live client contract (setup); no native Python server
Tool generation for agents	Available	Under consideration
Serve: auth, CORS, rate limiting, errors	Available	Available
Serve: JWT strategy and token issuer	Available	Available (`jwt_strategy`, `analytics_token_issuer`; tokens verify across languages)
Serve: your app's own login (`fromContext`)	Available	Available (`from_context`, `django_user`, `flask_session`)
Serve adapters	Fetch, Node, Vercel	ASGI, WSGI, native Django and Flask (`to_asgi_app`, `to_wsgi_app`, `urls`, `blueprint`)
Serve: dataset query endpoints and discovery	Available	Available
CLI (`init`, `generate datasets`, `dev`, `export client-contract`)	Available	Available
Generated React hooks	Available	Available via exported client contracts and `generate:client` (setup)
Query builder	Available	Not planned
Named metrics	Available	Not planned. Use metric endpoints or derived measures