> ## 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. # Predict (TabPFN JSON API) > Predict using uploaded test features and a fitted training set. Select classification or regression in `task_config`; its `predict_params` controls the output. The response contains predictions, model metadata, and optional stage timings. Long-running requests may include keepalive whitespace before the final JSON payload. ## OpenAPI ````yaml /api-reference/openapi.json post /tabpfn/predict openapi: 3.1.0 info: title: TabPFN API description: >- TabPFN REST API. For Python applications, use [tabpfn-client](/api-reference/overview). See the [REST quickstart](/api/rest-quickstart) for authentication and the upload, fit, and predict flow. version: 2.0.0 contact: name: Prior Labs email: hello@priorlabs.ai servers: - url: https://api.priorlabs.ai description: Production TabPFN API security: - BearerAuth: [] paths: /tabpfn/predict: post: tags: - Prediction summary: Predict (TabPFN JSON API) description: >- Predict using uploaded test features and a fitted training set. Select classification or regression in `task_config`; its `predict_params` controls the output. The response contains predictions, model metadata, and optional stage timings. Long-running requests may include keepalive whitespace before the final JSON payload. operationId: tabpfn_predict requestBody: required: true content: application/json: schema: $ref: '#/components/schemas/PredictRequest' example: test_set_upload_id: 33333333-3333-4333-8333-333333333333 fitted_train_set_id: 22222222-2222-4222-8222-222222222222 task_config: task: classification predict_params: output_type: probas responses: '200': description: Successful response content: application/json: schema: $ref: '#/components/schemas/PredictResponse' default: description: Error response content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' components: schemas: PredictRequest: additionalProperties: false properties: test_set_upload_id: format: uuid title: Test Set Upload Id type: string fitted_train_set_id: format: uuid title: Fitted Train Set Id type: string task_config: discriminator: mapping: classification: '#/components/schemas/ClassifierConfig' regression: '#/components/schemas/RegressorConfig' propertyName: task oneOf: - $ref: '#/components/schemas/ClassifierConfig' - $ref: '#/components/schemas/RegressorConfig' title: Task Config required: - test_set_upload_id - fitted_train_set_id - task_config title: PredictRequest type: object PredictResponse: additionalProperties: false properties: prediction: anyOf: - items: {} type: array - items: items: {} type: array type: array - additionalProperties: anyOf: - items: {} type: array - items: items: {} type: array type: array type: object title: Prediction metadata: discriminator: mapping: classification: '#/components/schemas/ClassifierMetadata' regression: '#/components/schemas/RegressorMetadata' propertyName: task oneOf: - $ref: '#/components/schemas/ClassifierMetadata' - $ref: '#/components/schemas/RegressorMetadata' title: Metadata timings: anyOf: - $ref: '#/components/schemas/PredictTimings' - type: 'null' default: null required: - prediction - metadata title: PredictResponse type: object ErrorResponse: additionalProperties: false properties: message: title: Message type: string error_code: default: GENERAL_ERROR title: Error Code type: string x-sdk-keep-default: true trace_id: anyOf: - format: uuid type: string - type: 'null' default: null title: Trace Id required: - message title: ErrorResponse type: object ClassifierConfig: additionalProperties: false properties: task: const: classification default: classification title: Task type: string tabpfn_config: $ref: '#/components/schemas/ClassifierTabPFNConfig' predict_params: $ref: '#/components/schemas/ClassifierPredictParams' title: ClassifierConfig type: object RegressorConfig: additionalProperties: false properties: task: const: regression default: regression title: Task type: string tabpfn_config: $ref: '#/components/schemas/RegressorTabPFNConfig' predict_params: $ref: '#/components/schemas/RegressorPredictParams' title: RegressorConfig type: object ClassifierMetadata: additionalProperties: false properties: test_set_num_rows: title: Test Set Num Rows type: integer test_set_num_cols: title: Test Set Num Cols type: integer n_estimators: anyOf: - type: integer - type: 'null' default: null title: N Estimators billing_model_version: anyOf: - type: string - type: 'null' default: null title: Billing Model Version execution_mode: anyOf: - enum: - standard - thinking - cache type: string - type: 'null' default: null title: Execution Mode cache_outcome: anyOf: - enum: - not_requested - hit - miss - fallback type: string - type: 'null' default: null title: Cache Outcome task: const: classification default: classification title: Task type: string package_version: title: Package Version type: string tabpfn_config: $ref: '#/components/schemas/ClassifierTabPFNConfig' classes: anyOf: - items: anyOf: - type: string - type: integer - type: number - type: boolean type: array - type: 'null' default: null title: Classes top_k: anyOf: - type: integer - type: 'null' default: null title: Top K required: - test_set_num_rows - test_set_num_cols - package_version - tabpfn_config title: ClassifierMetadata type: object RegressorMetadata: additionalProperties: false properties: test_set_num_rows: title: Test Set Num Rows type: integer test_set_num_cols: title: Test Set Num Cols type: integer n_estimators: anyOf: - type: integer - type: 'null' default: null title: N Estimators billing_model_version: anyOf: - type: string - type: 'null' default: null title: Billing Model Version execution_mode: anyOf: - enum: - standard - thinking - cache type: string - type: 'null' default: null title: Execution Mode cache_outcome: anyOf: - enum: - not_requested - hit - miss - fallback type: string - type: 'null' default: null title: Cache Outcome task: const: regression default: regression title: Task type: string package_version: title: Package Version type: string tabpfn_config: $ref: '#/components/schemas/RegressorTabPFNConfig' required: - test_set_num_rows - test_set_num_cols - package_version - tabpfn_config title: RegressorMetadata type: object PredictTimings: additionalProperties: false description: >- Where a predict's time went, per stage. Queue wait covers everything outside worker processing: waiting for a free worker and transport. A refit the server triggers counts towards the stage it happened in. properties: test_set_transform_queue_wait_s: anyOf: - type: number - type: 'null' default: null description: >- Seconds the test set transform waited for a worker. Null when an earlier request already transformed this test set. title: Test Set Transform Queue Wait S test_set_transform_s: anyOf: - type: number - type: 'null' default: null description: >- Seconds a worker spent loading, validating and preprocessing the test set. Null when an earlier request already transformed this test set. title: Test Set Transform S predict_queue_wait_s: anyOf: - type: number - type: 'null' default: null description: Seconds the prediction waited for a worker. title: Predict Queue Wait S predict_s: anyOf: - type: number - type: 'null' default: null description: Seconds a worker spent predicting. title: Predict S title: PredictTimings type: object ClassifierTabPFNConfig: additionalProperties: false properties: n_estimators: anyOf: - maximum: 8 minimum: 1 type: integer - type: 'null' default: null title: N Estimators categorical_features_indices: anyOf: - items: type: integer type: array - type: 'null' default: null title: Categorical Features Indices softmax_temperature: anyOf: - type: number - type: 'null' default: null title: Softmax Temperature average_before_softmax: anyOf: - type: boolean - type: 'null' default: null title: Average Before Softmax random_state: anyOf: - type: integer - type: 'null' default: null title: Random State inference_config: anyOf: - additionalProperties: true type: object - type: 'null' default: null description: Refer to tabpfn.inference_config.InferenceConfig for more details. title: Inference Config inference_precision: anyOf: - enum: - autocast - auto type: string - type: 'null' default: null title: Inference Precision ignore_pretraining_limits: default: true title: Ignore Pretraining Limits type: boolean n_preprocessing_jobs: default: 4 title: N Preprocessing Jobs type: integer fit_mode: $ref: '#/components/schemas/FitMode' default: fit_preprocessors device: anyOf: - items: type: string minItems: 1 type: array - type: 'null' default: null description: Stringified `torch.device`s title: Device memory_saving_mode: anyOf: - type: boolean - type: 'null' default: null title: Memory Saving Mode model_path: anyOf: - type: string - type: 'null' default: null title: Model Path balance_probabilities: anyOf: - type: boolean - type: 'null' default: null title: Balance Probabilities title: ClassifierTabPFNConfig type: object ClassifierPredictParams: additionalProperties: false properties: output_type: $ref: '#/components/schemas/ClassifierOutputType' default: probas top_k: anyOf: - minimum: 1 type: integer - type: 'null' default: null title: Top K title: ClassifierPredictParams type: object RegressorTabPFNConfig: additionalProperties: false properties: n_estimators: anyOf: - maximum: 8 minimum: 1 type: integer - type: 'null' default: null title: N Estimators categorical_features_indices: anyOf: - items: type: integer type: array - type: 'null' default: null title: Categorical Features Indices softmax_temperature: anyOf: - type: number - type: 'null' default: null title: Softmax Temperature average_before_softmax: anyOf: - type: boolean - type: 'null' default: null title: Average Before Softmax random_state: anyOf: - type: integer - type: 'null' default: null title: Random State inference_config: anyOf: - additionalProperties: true type: object - type: 'null' default: null description: Refer to tabpfn.inference_config.InferenceConfig for more details. title: Inference Config inference_precision: anyOf: - enum: - autocast - auto type: string - type: 'null' default: null title: Inference Precision ignore_pretraining_limits: default: true title: Ignore Pretraining Limits type: boolean n_preprocessing_jobs: default: 4 title: N Preprocessing Jobs type: integer fit_mode: $ref: '#/components/schemas/FitMode' default: fit_preprocessors device: anyOf: - items: type: string minItems: 1 type: array - type: 'null' default: null description: Stringified `torch.device`s title: Device memory_saving_mode: anyOf: - type: boolean - type: 'null' default: null title: Memory Saving Mode model_path: anyOf: - type: string - type: 'null' default: null title: Model Path title: RegressorTabPFNConfig type: object RegressorPredictParams: additionalProperties: false properties: output_type: $ref: '#/components/schemas/RegressorOutputType' default: mean quantiles: anyOf: - items: maximum: 1 minimum: 0 type: number type: array - type: 'null' default: null title: Quantiles title: RegressorPredictParams type: object FitMode: enum: - low_memory - fit_preprocessors - fit_with_cache - batched title: FitMode type: string ClassifierOutputType: enum: - probas - preds - top_k title: ClassifierOutputType type: string RegressorOutputType: enum: - mean - median - mode - quantiles - full - main title: RegressorOutputType type: string securitySchemes: BearerAuth: type: http scheme: bearer bearerFormat: JWT description: >- Bearer token for authentication, obtained after signing up and generating an API key. ```` This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.