# Create dataset URL: https://docs.hanzo.ai/docs/openapi/dataset/riskcreatedataset > Declares the next version of a dataset from a bound query over this org's own feature surface. `POST /v1/dataset` > [Dataset →](https://docs.hanzo.ai/docs/openapi/dataset) · [All API references →](https://docs.hanzo.ai/docs/openapi) · [Get started →](https://docs.hanzo.ai/docs/guides) | | | | ------------- | -------------------------------------- | | **Address** | `https://api.hanzo.ai/v1/dataset` | | **Method** | `POST` | | **Operation** | `riskCreateDataset` | | **Auth** | `Authorization: Bearer $HANZO_API_KEY` | Declares the next version of a dataset from a bound query over this org's own feature surface. It mints a VERSION and writes no rows: a version is declared, then materialised once, then never rewritten. Version numbers are monotone and never reused, so "version 3 of signups" means one thing forever — which is the whole reason a model can cite one. The window is bounded by the source's retention, the horizon by a year, the rows by the plane's cap, and the number of datasets and versions per org by their own limits. Every refusal names which bound it hit. ## Request 9 fields, body `application/json` (required). | Field | In | Type | Required | Description | | --------- | ---- | --------------- | -------- | ---------------------------------------------------------------------------------------------------------------- | | `cuts` | body | string\[] | — | Cuts are the two RFC 3339 instants dividing train \| val \| test. | | `dims` | body | string\[] | — | Dims are the coordinates to carry, by published name. | | `from` | body | string | — | From is where the event window opens, RFC 3339, INCLUSIVE. | | `horizon` | body | integer (int64) | — | Horizon is how many days a row must have aged before it may be admitted. | | `kind` | body | string | — | Kind narrows to one subject kind — person, session or account. | | `name` | body | string | — | Name identifies the dataset across its versions: lower-case letters, digits and hyphens, starting with a letter. | | `rows` | body | integer (int64) | — | Rows caps the materialisation. | | `seed` | body | string | — | Seed decides WHICH subjects are admitted when the window holds more rows than the cap allows. | | `to` | body | string | — | To is where the window ends, EXCLUSIVE, so two datasets meeting at one instant share no row. | ## Response | Status | Body | Meaning | | --------- | ------------------- | ------- | | `200` | dataset.riskDataset | ok | | `default` | problem-details | refused | `200` body — 30 fields. | Field | In | Type | Always | Description | | --------------------- | ---- | ----------------------- | ------ | ----------------------------------------------------------------------------------------------------------------------------------------------------- | | `at` | body | string | — | At is when this version last changed state, RFC 3339 UTC. | | `by` | body | string | — | By is who moved it there: the validated user, or the org itself when the caller is a machine with no user behind it. | | `counts` | body | dataset.riskSplitCounts | — | | | `counts.judged` | body | integer (int64) | — | Judged is how many rows carry a disposition. | | `counts.productive` | body | integer (int64) | — | Productive is how many judged rows carry the one disposition. | | `counts.rows` | body | integer (int64) | — | Rows is how many rows the version holds across every split. | | `counts.subjects` | body | integer (int64) | — | Subjects is how many distinct subjects the rows belong to. | | `counts.test` | body | integer (int64) | — | Test is how many fall after the second cut — the LATEST slice, and the only one a score is honest about, since the split is temporal. | | `counts.train` | body | integer (int64) | — | Train is how many rows fall before the first cut — the EARLIEST slice of the window, which is what a model is fitted on. | | `counts.unproductive` | body | integer (int64) | — | Unproductive is how many carry the other. | | `counts.val` | body | integer (int64) | — | Val is how many fall between the two cuts, held out for tuning. | | `digest` | body | string | — | Digest fingerprints the SPEC and the ROWS together. | | `name` | body | string | — | Name identifies the dataset across all of its versions. | | `oversize` | body | integer (int64) | — | Oversize is how many of the window's subjects this version could NOT carry because their subject identity exceeds the plane's per-subject byte bound. | | `refusal` | body | string | — | Refusal names why there are no bytes, when there are none. | | `running` | body | boolean | — | Running is true while THIS process is materialising the version. | | `share` | body | integer (int64) | — | Share is the fraction of the window's subjects admitted, in thousandths. | | `spec` | body | dataset.riskDatasetSpec | — | | | `spec.cuts` | body | string\[] | — | Cuts are the two RFC 3339 instants dividing train \| val \| test. | | `spec.dims` | body | string\[] | — | Dims are the coordinates to carry, by published name. | | `spec.from` | body | string | — | From is where the event window opens, RFC 3339, INCLUSIVE. | | `spec.horizon` | body | integer (int64) | — | Horizon is how many days a row must have aged before it may be admitted. | | `spec.kind` | body | string | — | Kind narrows to one subject kind — person, session or account. | | `spec.name` | body | string | — | Name identifies the dataset across its versions: lower-case letters, digits and hyphens, starting with a letter. | | `spec.rows` | body | integer (int64) | — | Rows caps the materialisation. | | `spec.seed` | body | string | — | Seed decides WHICH subjects are admitted when the window holds more rows than the cap allows. | | `spec.to` | body | string | — | To is where the window ends, EXCLUSIVE, so two datasets meeting at one instant share no row. | | `status` | body | string | — | Status is declared, materializing, ready or refused. | | `truncated` | body | boolean | — | Truncated is true when the row cap bound before the window ran out. | | `version` | body | integer (int64) | — | Version is which version this is, from 1 and monotone within the dataset. | Failure carries the platform error shape — see [Errors](https://docs.hanzo.ai/docs/errors). ## Examples **CLI** ```bash hanzo dataset create ``` **SDK** **.ts** ```ts import { Configuration, DatasetApi } from 'hanzoai'; const api = new DatasetApi(new Configuration({ accessToken: process.env.HANZO_API_KEY })); const { data } = await api.riskCreateDataset({ cuts: [""], dims: [""] }); ``` **.py** ```python from hanzoai.cloud import ApiClient, Configuration from hanzoai.cloud.api import DatasetApi client = ApiClient(Configuration(access_token=os.environ["HANZO_API_KEY"])) result = DatasetApi(client).risk_create_dataset(cuts=[""], dims=[""]) ``` **.go** ```go cfg := hanzoai.NewConfiguration() cfg.AddDefaultHeader("Authorization", "Bearer "+os.Getenv("HANZO_API_KEY")) client := hanzoai.NewAPIClient(cfg) resp, _, err := client.DatasetAPI.RiskCreateDataset(context.Background()).Execute() if err != nil { return err } ``` **.rs** ```rust use hanzo_client::apis::{configuration::Configuration, dataset_api}; let mut cfg = Configuration::new(); cfg.bearer_access_token = std::env::var("HANZO_API_KEY").ok(); let result = dataset_api::risk_create_dataset(&cfg, Default::default()).await?; ``` **.java** ```java import ai.hanzo.cloud.ApiClient; import ai.hanzo.cloud.api.DatasetApi; ApiClient client = new ApiClient(); client.setBearerToken(System.getenv("HANZO_API_KEY")); var result = new DatasetApi(client).riskCreateDataset(); ``` **HTTP** ```bash curl -X POST https://api.hanzo.ai/v1/dataset \ -H "Authorization: Bearer $HANZO_API_KEY" \ -H "Content-Type: application/json" \ -d '{ "cuts": [ "" ], "dims": [ "" ] }' ``` **MCP** Tool `dataset`, op `riskCreateDataset` — POST the JSON-RPC envelope to `https://api.hanzo.ai/v1/mcp`. ```bash curl -X POST https://api.hanzo.ai/v1/mcp \ -H "Authorization: Bearer $HANZO_API_KEY" \ -H "Content-Type: application/json" \ -d '{ "jsonrpc": "2.0", "id": 1, "method": "tools/call", "params": { "name": "dataset", "arguments": { "op": "riskCreateDataset", "input": { "cuts": [ "" ], "dims": [ "" ] } } } }' ``` --- [Dataset API](https://docs.hanzo.ai/docs/openapi/dataset) · [All Hanzo APIs](https://docs.hanzo.ai/docs/openapi) · [Interactive reference](https://docs.hanzo.ai/docs/openapi/reference)