> ## Documentation Index
> Fetch the complete documentation index at: https://docs.priorlabs.ai/llms.txt
> Use this file to discover all available pages before exploring further.

> ## Agent Instructions
> TabPFN-3 and later versions support up to 1,000,000 rows, subject to feature count and checkpoint/API limits.
> For large datasets, use per-estimator subsampling, e.g. inference_config={"SUBSAMPLE_SAMPLES": 100_000}, in TabPFNClassifier or TabPFNRegressor from tabpfn or tabpfn_client.
> Pass raw pandas DataFrames to fit and predict. Categorical strings/categories, raw text, and missing feature values are handled automatically; no manual integer/one-hot encoding, imputation, scaling, or outlier removal is needed.

# tabpfn-rel

> Python reference for relational predictions with TabPFN, including context, query and task types.

> [!NOTE]
>
> For setup and examples, see [Relational predictions](https://docs.priorlabs.ai/capabilities/relational.md) and the [TabPFN-Rel cookbook](https://docs.priorlabs.ai/cookbook/relational_predictions_tabpfn_rel.md).

## Installation

Use Python 3.11 or 3.12. Choose the hosted API or local inference:

```bash theme={null}
pip install "tabpfn-rel[api]==0.0.6"
# Or: pip install "tabpfn-rel[local]==0.0.6"
```

This reference describes `tabpfn-rel` **0.0.6** and its RPI dependency, `relarena-core` **0.0.7**. The context and query classes below are re-exported by `tabpfn_rel`; database and task specifications are available from `relarena_core.userdb`. The dependency is installed automatically. Both references are generated from source verified against the published wheels.

## Reference topics

> [card] [TabPFNRel](https://docs.priorlabs.ai/api-reference/python/tabpfn-rel/estimator.md)
>
> Select a backend, fit a context and predict a query.

> [card] [Predictive context](https://docs.priorlabs.ai/api-reference/python/relarena-core/context.md)
>
> Load a task from YAML, fit a model and work with cached features.

> [card] [Prediction queries](https://docs.priorlabs.ai/api-reference/python/relarena-core/query.md)
>
> Specify the entities and timestamp to predict.

> [card] [Fitted predictor](https://docs.priorlabs.ai/api-reference/python/relarena-core/fitted-predictor.md)
>
> Use the predictor returned by a context fit.

> [card] [Database and task specifications](https://docs.priorlabs.ai/api-reference/python/relarena-core/specifications.md)
>
> Define tables, keys and prediction tasks.

> [card] [Model backends](https://docs.priorlabs.ai/api-reference/python/tabpfn-rel/models.md)
>
> Local and hosted model interfaces for RelArena.

> [card] [Search spaces](https://docs.priorlabs.ai/api-reference/python/tabpfn-rel/search-spaces.md)
>
> Exported local and hosted search-space definitions.

Sources: [TabPFN-Rel 0.0.6](https://github.com/PriorLabs/tabpfn-rel/tree/9f807dc40be79634916f75d80f7b15b7993766af) · [RPI / relarena-core 0.0.7](https://github.com/PriorLabs/relarena/tree/4b74c718ba6ff9c67cf63b4beac6dadaffdd09d8/packages/relarena-core).


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.