# Toloka Platform

> Label your data, train a model, and build a repeatable workflow on Toloka Platform.

- [Get started with Toloka Platform](https://docs.platform.toloka.ai/docs/index.md): Label your data, train a model, and build a repeatable workflow.
- [How the platform works](https://docs.platform.toloka.ai/docs/start/how-it-works.md): Choose Data Labeling or Train, then work with shared projects, datasets, and a team balance.
- [Glossary](https://docs.platform.toloka.ai/docs/start/glossary.md): Shared Platform terms, Data Labeling vocabulary, and Train concepts.
- [Get your first labeled dataset](https://docs.platform.toloka.ai/docs/data/get-started/first-labeled-dataset.md): Label 10 support tickets by intent and download the results.
- [How data labeling works](https://docs.platform.toloka.ai/docs/data/concepts/overview.md): How work is divided between you and the agent — you describe the task and approve the calls that cost money; the agent builds, configures, and checks the pipeline.
- [Projects](https://docs.platform.toloka.ai/docs/core/projects.md): Use projects as the shared workspace for datasets, Data Labeling pipelines, Training drafts, models, files, and secrets.
- [Create and manage datasets](https://docs.platform.toloka.ai/docs/core/datasets.md): Create and manage project datasets shared by Data Labeling and Train, including upload rules, fields, filters, exports, and Training conversation rows.
- [Pipelines](https://docs.platform.toloka.ai/docs/data/concepts/pipelines.md): Choose nodes for expert or synthetic labeling, multiple answers, routing, and computation.
- [Fields and data](https://docs.platform.toloka.ai/docs/data/concepts/fields-and-data.md): How pipelines read and write dataset fields, including required, optional, repeating, and per-expert values.
- [Quality control](https://docs.platform.toloka.ai/docs/data/concepts/quality-control.md): How the platform checks labeled data — the quality criteria each output is judged against, who reviews the work, the verdicts you get back, what a failed check triggers, and what review costs.
- [Self-check](https://docs.platform.toloka.ai/docs/data/concepts/self-check.md): Validate node settings, run checks, and resolve errors or warnings before deployment.
- [Pipeline pricing](https://docs.platform.toloka.ai/docs/data/concepts/pricing.md): How Data Labeling pipeline runs are forecast, which cost lines apply, what Expected and Maximum mean, how deployment freezes a contract, and what canceled or failed items cost.
- [Review the labeling interface](https://docs.platform.toloka.ai/docs/data/guides/review-labeling-interface.md): Preview and edit the labeling interface, then test a submission with Annotation Quality Assurance.
- [Label a large dataset in batches](https://docs.platform.toloka.ai/docs/data/guides/label-at-volume.md): Label more than one run's worth of data, and repeat the cycle for a recurring workload, by deploying in batches of up to 10,000 items and cloning the pipeline for each new batch.
- [Label with an LLM instead of experts](https://docs.platform.toloka.ai/docs/data/guides/synthetic-labeling.md): Have an LLM label a node's data instead of expert labelers, with each label checked against your quality criteria by a separate LLM call.
- [Export your labeled results](https://docs.platform.toloka.ai/docs/data/guides/export-results.md): Export the results of a pipeline run as a JSON file and download it from the Items tab.
- [Resume or cancel a pipeline run](https://docs.platform.toloka.ai/docs/data/guides/pause-resume-cancel.md): Resume a run after an automatic funding pause, or cancel it and understand the effects on results and charges.
- [Start a project from a template](https://docs.platform.toloka.ai/docs/data/recipes/preset-project.md): Launch a project from a ready-made template — pick what you need done, open a template, answer the assistant's questions, attach your data, and deploy.
- [Have the agent build your project](https://docs.platform.toloka.ai/docs/data/recipes/agentic-project.md): Describe the data you need in plain language, review the plan the agent drafts, and let it build the project before you deploy.
- [Pipeline nodes](https://docs.platform.toloka.ai/docs/data/reference/nodes/index.md): The kinds of nodes a pipeline can contain, what each one does, and the fields and options each exposes.
- [Start and End nodes](https://docs.platform.toloka.ai/docs/data/reference/nodes/start-and-end.md): The pipeline's entry and exit — the dataset the Start node reads, how its filter and item limit choose which items a run launches, which dataset fields become inputs, and which produced fields the End node returns.
- [Generation node](https://docs.platform.toloka.ai/docs/data/reference/nodes/generation.md): Expert and synthetic labeling settings, fields, quality control, onboarding, and pricing.
- [Quorum node](https://docs.platform.toloka.ai/docs/data/reference/nodes/quorum.md): The Quorum node in full — how many experts label each item, the array shape of every output, quality control by LLM review or expert agreement, the expert audience, and pricing.
- [Code node](https://docs.platform.toloka.ai/docs/data/reference/nodes/code.md): The Code node in full — the fields it reads and writes, the async main function it runs, the runtime environment, environment variables and secrets, the editing check cases, and the run-time limits and failures.
- [Match node](https://docs.platform.toloka.ai/docs/data/reference/nodes/match.md): The Match node in full — the input fields its conditions read, how conditions combine field tests with all or any, the equal and not-equal text comparison, the matched and default branches, error routing, and how the node shapes the cost estimate.
- [Export format](https://docs.platform.toloka.ai/docs/data/reference/export-format.md): The export file in full — the JSON array a run's export produces, the fields on each item entry, the six status values, and the per-node metrics block.
- [Choose a training method](https://docs.platform.toloka.ai/docs/train/get-started/choose-method.md): Compare supervised fine-tuning and prompt gisting by input, output, model support, and intended use.
- [Train your first model](https://docs.platform.toloka.ai/docs/train/get-started/first-trained-model.md): Split 10 support tickets into train, validation, and test parts, then train and download a Model Version.
- [How Training works](https://docs.platform.toloka.ai/docs/train/concepts/how-training-works.md): Understand Training methods, data, lifecycle, run states, outputs, and supported product surfaces.
- [Training cost](https://docs.platform.toloka.ai/docs/train/concepts/cost.md): Understand Training cost estimates, maximum charges, balance holds, settlement, and no-charge cases.
- [Prepare training data from labels](https://docs.platform.toloka.ai/docs/train/guides/prepare-training-data.md): Convert a Data Labeling export into chat training rows while keeping a deterministic evaluation set out of Training.
- [Fine-tune a model](https://docs.platform.toloka.ai/docs/train/guides/fine-tune.md): Sign in, load labeled conversations, create an SFT Training, review the cost, and start one Training Run.
- [Compress a prompt with gisting](https://docs.platform.toloka.ai/docs/train/guides/gist.md): Sign in, load examples, attach a prompt file, create a gisting Training, review the cost, and start one Training Run.
- [Base models](https://docs.platform.toloka.ai/docs/train/reference/base-models.md): Reference the Training model catalog, method compatibility, GPU counts, pinned revisions, and model lineage.
- [Training settings](https://docs.platform.toloka.ai/docs/train/reference/settings.md): Reference SFT and gisting settings, accepted values, defaults, cost effects, and CLI flags.
- [Training limits](https://docs.platform.toloka.ai/docs/train/reference/limits.md): Reference Training names, data size, prompt, estimate, runtime, Model Version, and download limits.
- [Label, train, and evaluate](https://docs.platform.toloka.ai/docs/workflows/label-train-evaluate.md): Turn completed Data Labeling results into a fine-tuning dataset, keep a golden set separate, and score predictions from your trained model locally.
- [Teams, access, and balance](https://docs.platform.toloka.ai/docs/core/teams-and-balance.md): Manage team access in the Business Console, switch teams, add funds or promocodes, and view billing history.
- [Billing](https://docs.platform.toloka.ai/docs/core/billing.md): How Data Labeling pipeline runs and Training Runs share your team's balance, reserve funds, charge for work, and respond to insufficient funds.
- [Security and privacy](https://docs.platform.toloka.ai/docs/core/security.md): Find Toloka's security policies, privacy terms, and the right contact for a data-handling question.
- [TypeScript SDK](https://docs.platform.toloka.ai/docs/sdk/index.md): Call the HTTP API, define pipelines as code, and check your project locally.
- [SDK quickstart](https://docs.platform.toloka.ai/docs/sdk/quickstart.md): Create a script that authenticates, adds dataset rows, and exports them.
- [Client and credentials](https://docs.platform.toloka.ai/docs/sdk/client.md): Configure the client, issue project-scoped tokens, and handle request limits.
- [Pagination](https://docs.platform.toloka.ai/docs/sdk/pagination.md): Fetch one page at a time or iterate through all matching resources.
- [Long-running jobs](https://docs.platform.toloka.ai/docs/sdk/long-running-jobs.md): Wait for background work and distinguish completion, failure, and timeouts.
- [Errors](https://docs.platform.toloka.ai/docs/sdk/errors.md): Handle HTTP errors, wait timeouts, and local pipeline validation failures.
- [Pipelines as code](https://docs.platform.toloka.ai/docs/sdk/pipelines.md): Define Data Labeling pipelines as TypeScript source, then check, plan, and apply them.
- [Local testing](https://docs.platform.toloka.ai/docs/sdk/testing.md): Run Python Code nodes and test labeling interfaces and pipeline flows with Vitest.
- [CLI](https://docs.platform.toloka.ai/docs/cli.md): Install and use the toloka CLI for shared authentication and project setup, Data Labeling pipelines, and Train workflows.
- [Reference](https://docs.platform.toloka.ai/docs/reference/index.md): SDK methods, HTTP endpoints, and CLI commands.
- [Client reference](https://docs.platform.toloka.ai/docs/sdk/reference/client-methods.md): Find SDK methods, signatures, return types, and HTTP endpoints by resource.
- [createClient](https://docs.platform.toloka.ai/docs/sdk/reference/create-client.md): Create a typed client for the platform's HTTP API.
- [waitFor](https://docs.platform.toloka.ai/docs/sdk/reference/wait-for.md): Poll a background operation until it reaches its completion condition.
- [client.me()](https://docs.platform.toloka.ai/docs/sdk/reference/account/me.md): Get the current user.
- [client.projects.listAll()](https://docs.platform.toloka.ai/docs/sdk/reference/projects/list.md): List projects.
- [client.projects.create()](https://docs.platform.toloka.ai/docs/sdk/reference/projects/create.md): Create a project.
- [client.projects.get()](https://docs.platform.toloka.ai/docs/sdk/reference/projects/get.md): Get a project.
- [client.projects.update()](https://docs.platform.toloka.ai/docs/sdk/reference/projects/update.md): Update a project.
- [client.projects.createToken()](https://docs.platform.toloka.ai/docs/sdk/reference/projects/create-token.md): Issue a project-scoped token.
- [client.projects.listAllTokens()](https://docs.platform.toloka.ai/docs/sdk/reference/projects/list-tokens.md): List live project-scoped tokens.
- [client.projects.revokeAllTokens()](https://docs.platform.toloka.ai/docs/sdk/reference/projects/revoke-all-tokens.md): Revoke all project tokens.
- [client.projects.revokeToken()](https://docs.platform.toloka.ai/docs/sdk/reference/projects/revoke-token.md): Revoke a project-scoped token.
- [client.projects.revokeCurrentToken()](https://docs.platform.toloka.ai/docs/sdk/reference/projects/revoke-current-token.md): Revoke the current token.
- [client.datasets.listAll()](https://docs.platform.toloka.ai/docs/sdk/reference/datasets/list.md): List project datasets.
- [client.datasets.create()](https://docs.platform.toloka.ai/docs/sdk/reference/datasets/create.md): Create a dataset.
- [client.datasets.get()](https://docs.platform.toloka.ai/docs/sdk/reference/datasets/get.md): Get a dataset.
- [client.datasets.listAllFields()](https://docs.platform.toloka.ai/docs/sdk/reference/datasets/list-fields.md): List dataset fields.
- [client.datasets.createField()](https://docs.platform.toloka.ai/docs/sdk/reference/datasets/create-field.md): Create dataset fields.
- [client.datasets.getInferredFields()](https://docs.platform.toloka.ai/docs/sdk/reference/datasets/get-inferred-fields.md): List inferred dataset fields.
- [client.datasets.appendItems()](https://docs.platform.toloka.ai/docs/sdk/reference/datasets/append-items.md): Add rows to a dataset.
- [client.datasets.listAllItems()](https://docs.platform.toloka.ai/docs/sdk/reference/datasets/list-items.md): List dataset items.
- [client.datasets.createSlice()](https://docs.platform.toloka.ai/docs/sdk/reference/datasets/create-slice.md): Create a dataset slice.
- [client.datasets.listSlicesPage()](https://docs.platform.toloka.ai/docs/sdk/reference/datasets/list-slices.md): List dataset slices.
- [client.datasets.getSlice()](https://docs.platform.toloka.ai/docs/sdk/reference/datasets/get-slice.md): Get a dataset slice.
- [client.datasets.split()](https://docs.platform.toloka.ai/docs/sdk/reference/datasets/split.md): Split a dataset into parts.
- [client.datasets.createFieldsFromUploadedRows()](https://docs.platform.toloka.ai/docs/sdk/reference/datasets/create-fields-from-uploaded-rows.md): Create fields from uploaded rows.
- [client.datasets.startUpload()](https://docs.platform.toloka.ai/docs/sdk/reference/datasets/start-upload.md): Start a dataset file upload.
- [client.datasets.finalizeUpload()](https://docs.platform.toloka.ai/docs/sdk/reference/datasets/finalize-upload.md): Finalize a dataset file upload.
- [client.datasets.cancelUpload()](https://docs.platform.toloka.ai/docs/sdk/reference/datasets/cancel-upload.md): Cancel a dataset upload.
- [client.datasets.getUpload()](https://docs.platform.toloka.ai/docs/sdk/reference/datasets/get-upload.md): Get a dataset upload.
- [client.datasets.startExport()](https://docs.platform.toloka.ai/docs/sdk/reference/datasets/start-export.md): Start a dataset export.
- [client.datasets.getExport()](https://docs.platform.toloka.ai/docs/sdk/reference/datasets/get-export.md): Get a dataset export.
- [client.pipelines.listAll()](https://docs.platform.toloka.ai/docs/sdk/reference/pipelines/list.md): List project pipelines.
- [client.pipelines.get()](https://docs.platform.toloka.ai/docs/sdk/reference/pipelines/get.md): Get a pipeline.
- [client.pipelines.update()](https://docs.platform.toloka.ai/docs/sdk/reference/pipelines/update.md): Update a pipeline.
- [client.pipelines.import()](https://docs.platform.toloka.ai/docs/sdk/reference/pipelines/import.md): Import a pipeline into a project.
- [client.pipelines.attachDataset()](https://docs.platform.toloka.ai/docs/sdk/reference/pipelines/attach-dataset.md): Attach a dataset to a pipeline.
- [client.pipelines.export()](https://docs.platform.toloka.ai/docs/sdk/reference/pipelines/export.md): Export a pipeline.
- [client.pipelines.copy()](https://docs.platform.toloka.ai/docs/sdk/reference/pipelines/copy.md): Copy a pipeline.
- [client.pipelines.estimateCost()](https://docs.platform.toloka.ai/docs/sdk/reference/pipelines/estimate-cost.md): Estimate pipeline cost.
- [client.pipelines.getGraph()](https://docs.platform.toloka.ai/docs/sdk/reference/pipelines/get-graph.md): Get the pipeline graph.
- [client.pipelines.replaceGraph()](https://docs.platform.toloka.ai/docs/sdk/reference/pipelines/replace-graph.md): Replace the pipeline graph.
- [client.pipelines.validateGraph()](https://docs.platform.toloka.ai/docs/sdk/reference/pipelines/validate-graph.md): Validate a pipeline graph.
- [client.pipelines.patchNode()](https://docs.platform.toloka.ai/docs/sdk/reference/pipelines/patch-node.md): Update a pipeline node.
- [client.runs.start()](https://docs.platform.toloka.ai/docs/sdk/reference/runs/start.md): Start a pipeline run.
- [client.runs.listAll()](https://docs.platform.toloka.ai/docs/sdk/reference/runs/list.md): List pipeline runs.
- [client.runs.get()](https://docs.platform.toloka.ai/docs/sdk/reference/runs/get.md): Get a pipeline run.
- [client.runs.cancel()](https://docs.platform.toloka.ai/docs/sdk/reference/runs/cancel.md): Cancel a pipeline run.
- [client.runs.listAllItems()](https://docs.platform.toloka.ai/docs/sdk/reference/runs/list-items.md): List run items.
- [client.runs.getItem()](https://docs.platform.toloka.ai/docs/sdk/reference/runs/get-item.md): Get a run item.
- [client.checks.list()](https://docs.platform.toloka.ai/docs/sdk/reference/checks/list.md): List pipeline checks.
- [client.checks.start()](https://docs.platform.toloka.ai/docs/sdk/reference/checks/start.md): Start a pipeline check.
- [client.checks.runLints()](https://docs.platform.toloka.ai/docs/sdk/reference/checks/run-lints.md): Run pipeline lints.
- [client.checks.get()](https://docs.platform.toloka.ai/docs/sdk/reference/checks/get.md): Get a pipeline check.
- [client.secrets.list()](https://docs.platform.toloka.ai/docs/sdk/reference/secrets/list.md): List project secret names.
- [client.secrets.set()](https://docs.platform.toloka.ai/docs/sdk/reference/secrets/set.md): Set a project secret.
- [client.files.list()](https://docs.platform.toloka.ai/docs/sdk/reference/files/list.md): List project files.
- [client.files.startUpload()](https://docs.platform.toloka.ai/docs/sdk/reference/files/start-upload.md): Start a project file upload.
- [client.files.finalizeUpload()](https://docs.platform.toloka.ai/docs/sdk/reference/files/finalize-upload.md): Finalize a project file upload.
- [client.files.delete()](https://docs.platform.toloka.ai/docs/sdk/reference/files/delete.md): Delete a project file.
- [client.generationModels.list()](https://docs.platform.toloka.ai/docs/sdk/reference/generation-models/list.md): List synthetic generation models.
- [client.audience.specializations.listAll()](https://docs.platform.toloka.ai/docs/sdk/reference/audience/list-specializations.md): List expert specializations.
- [client.expertLists.list()](https://docs.platform.toloka.ai/docs/sdk/reference/expert-lists/list.md): List expert lists.
- [client.expertLists.create()](https://docs.platform.toloka.ai/docs/sdk/reference/expert-lists/create.md): Create an expert list.
- [client.expertLists.get()](https://docs.platform.toloka.ai/docs/sdk/reference/expert-lists/get.md): Get an expert list.
- [client.expertLists.update()](https://docs.platform.toloka.ai/docs/sdk/reference/expert-lists/update.md): Rename, hide or show an expert list.
- [client.expertLists.listExpertsPage()](https://docs.platform.toloka.ai/docs/sdk/reference/expert-lists/list-experts.md): List the experts of an expert list.
- [client.expertLists.addExperts()](https://docs.platform.toloka.ai/docs/sdk/reference/expert-lists/add-experts.md): Add experts to an expert list.
- [client.expertLists.removeExperts()](https://docs.platform.toloka.ai/docs/sdk/reference/expert-lists/remove-experts.md): Remove experts from an expert list.
- [client.expertBanList.get()](https://docs.platform.toloka.ai/docs/sdk/reference/expert-ban-list/get.md): Get the expert ban list of a project.
- [client.expertBanList.listExpertsPage()](https://docs.platform.toloka.ai/docs/sdk/reference/expert-ban-list/list-experts.md): List the experts of the expert ban list.
- [client.expertBanList.addExperts()](https://docs.platform.toloka.ai/docs/sdk/reference/expert-ban-list/add-experts.md): Ban experts in a project.
- [client.expertBanList.removeExperts()](https://docs.platform.toloka.ai/docs/sdk/reference/expert-ban-list/remove-experts.md): Unban experts in a project.
- [client.baseModels.list()](https://docs.platform.toloka.ai/docs/sdk/reference/base-models/list.md): List base models.
- [client.baseModels.get()](https://docs.platform.toloka.ai/docs/sdk/reference/base-models/get.md): Get a base model.
- [client.trainings.create()](https://docs.platform.toloka.ai/docs/sdk/reference/trainings/create.md): Create a training.
- [client.trainings.get()](https://docs.platform.toloka.ai/docs/sdk/reference/trainings/get.md): Get a training.
- [client.trainings.listAll()](https://docs.platform.toloka.ai/docs/sdk/reference/trainings/list.md): List project trainings.
- [client.trainings.updateMetadata()](https://docs.platform.toloka.ai/docs/sdk/reference/trainings/update-metadata.md): Update training metadata.
- [client.trainings.replaceConfiguration()](https://docs.platform.toloka.ai/docs/sdk/reference/trainings/replace-configuration.md): Replace a training configuration.
- [client.trainings.prepareCostEstimateInputs()](https://docs.platform.toloka.ai/docs/sdk/reference/trainings/prepare-cost-estimate-inputs.md): Prepare training cost estimate inputs.
- [client.trainings.estimateCost()](https://docs.platform.toloka.ai/docs/sdk/reference/trainings/estimate-cost.md): Estimate training cost.
- [client.trainings.createRun()](https://docs.platform.toloka.ai/docs/sdk/reference/trainings/create-run.md): Create a training run.
- [client.trainings.getRun()](https://docs.platform.toloka.ai/docs/sdk/reference/trainings/get-run.md): Get a training run.
- [client.trainings.listAllRuns()](https://docs.platform.toloka.ai/docs/sdk/reference/trainings/list-runs.md): List training runs.
- [client.trainings.getRunLogs()](https://docs.platform.toloka.ai/docs/sdk/reference/trainings/get-run-logs.md): Read training run logs.
- [client.trainings.getRunMetrics()](https://docs.platform.toloka.ai/docs/sdk/reference/trainings/get-run-metrics.md): Get training run metrics.
- [client.trainings.getRunMetricHistory()](https://docs.platform.toloka.ai/docs/sdk/reference/trainings/get-run-metric-history.md): Get training metric history.
- [client.trainings.cancelRun()](https://docs.platform.toloka.ai/docs/sdk/reference/trainings/cancel-run.md): Cancel a training run.
- [client.trainings.resumeRun()](https://docs.platform.toloka.ai/docs/sdk/reference/trainings/resume-run.md): Resume a training run.
- [client.models.listAll()](https://docs.platform.toloka.ai/docs/sdk/reference/models/list.md): List project models.
- [client.models.create()](https://docs.platform.toloka.ai/docs/sdk/reference/models/create.md): Create a project model.
- [client.models.listAllVersions()](https://docs.platform.toloka.ai/docs/sdk/reference/models/list-versions.md): List model versions.
- [client.models.getVersion()](https://docs.platform.toloka.ai/docs/sdk/reference/models/get-version.md): Get a model version.
- [client.models.getVersionContent()](https://docs.platform.toloka.ai/docs/sdk/reference/models/get-version-content.md): Get model version download links.
- [client.models.archive()](https://docs.platform.toloka.ai/docs/sdk/reference/models/archive.md): Archive a project model.
- [client.models.get()](https://docs.platform.toloka.ai/docs/sdk/reference/models/get.md): Get a project model.
- [client.batchInference.createRun()](https://docs.platform.toloka.ai/docs/sdk/reference/batch-inference/create-run.md): Create a batch inference run.
- [client.batchInference.listAllRuns()](https://docs.platform.toloka.ai/docs/sdk/reference/batch-inference/list-runs.md): List batch inference runs.
- [client.batchInference.getRun()](https://docs.platform.toloka.ai/docs/sdk/reference/batch-inference/get-run.md): Get a batch inference run.
- [client.batchInference.confirmRun()](https://docs.platform.toloka.ai/docs/sdk/reference/batch-inference/confirm-run.md): Confirm a batch inference run.
- [client.batchInference.resumeRun()](https://docs.platform.toloka.ai/docs/sdk/reference/batch-inference/resume-run.md): Resume a batch inference run.
- [client.batchInference.cancelRun()](https://docs.platform.toloka.ai/docs/sdk/reference/batch-inference/cancel-run.md): Cancel a batch inference run.
- [definePipeline](https://docs.platform.toloka.ai/docs/sdk/reference/define-pipeline.md): Build a pipeline definition from declared fields, nodes, and connections.
- [field](https://docs.platform.toloka.ai/docs/sdk/reference/field.md): Declare a named field with a JSON Schema and mark optional bindings with optional().
- [connect](https://docs.platform.toloka.ai/docs/sdk/reference/connect.md): Connect an output port to an input port in an authored pipeline.
- [root](https://docs.platform.toloka.ai/docs/sdk/reference/root.md): Define the Start and End field bindings of a pipeline.
- [code](https://docs.platform.toloka.ai/docs/sdk/reference/code.md): Create an authored Code node with field bindings and Python source.
- [generation](https://docs.platform.toloka.ai/docs/sdk/reference/generation.md): Create an authored Generation node using field objects in its entity settings.
- [quorum](https://docs.platform.toloka.ai/docs/sdk/reference/quorum.md): Create an authored Quorum node for collecting multiple annotations.
- [match](https://docs.platform.toloka.ai/docs/sdk/reference/match.md): Create an authored Match node with named conditions and branch ports.
- [validate](https://docs.platform.toloka.ai/docs/sdk/reference/validate.md): Assert that a value satisfies the SDK's pipeline graph contract.
- [CLI reference](https://docs.platform.toloka.ai/docs/cli/reference/index.md): Every public toloka command, argument, and option.
- [toloka version](https://docs.platform.toloka.ai/docs/cli/reference/version.md): Print the current CLI version.
- [toloka config](https://docs.platform.toloka.ai/docs/cli/reference/config.md): Read and write machine CLI settings.
- [toloka config get](https://docs.platform.toloka.ai/docs/cli/reference/config/get.md): Print a CLI setting, or every setting.
- [toloka config set](https://docs.platform.toloka.ai/docs/cli/reference/config/set.md): Write a CLI setting for this machine.
- [toloka login](https://docs.platform.toloka.ai/docs/cli/reference/login.md): Sign in through your browser, or store an API key, after verifying access.
- [toloka logout](https://docs.platform.toloka.ai/docs/cli/reference/logout.md): Remove stored credentials from this machine and end the browser login.
- [toloka auth-status](https://docs.platform.toloka.ai/docs/cli/reference/auth-status.md): Show which platform and credential the CLI would use.
- [toloka project](https://docs.platform.toloka.ai/docs/cli/reference/project.md): Select and manage the current platform project.
- [toloka project list](https://docs.platform.toloka.ai/docs/cli/reference/project/list.md): List remote projects this key can see.
- [toloka project add](https://docs.platform.toloka.ai/docs/cli/reference/project/add.md): Register a live project and select it.
- [toloka project switch](https://docs.platform.toloka.ai/docs/cli/reference/project/switch.md): Select a registered project.
- [toloka project current](https://docs.platform.toloka.ai/docs/cli/reference/project/current.md): Show the project this command would use.
- [toloka project url](https://docs.platform.toloka.ai/docs/cli/reference/project/url.md): Print the current project URL.
- [toloka project remove](https://docs.platform.toloka.ai/docs/cli/reference/project/remove.md): Forget a local project alias.
- [toloka project untrack](https://docs.platform.toloka.ai/docs/cli/reference/project/untrack.md): Remove a pipeline or dataset from local tracking without deleting it.
- [toloka project token](https://docs.platform.toloka.ai/docs/cli/reference/project/token.md): Mint a project-scoped token for agents and scripts; printed once, never stored.
- [toloka project tokens](https://docs.platform.toloka.ai/docs/cli/reference/project/tokens.md): List live project-scoped tokens.
- [toloka project revoke-token](https://docs.platform.toloka.ai/docs/cli/reference/project/revoke-token.md): Revoke a project-scoped token; it stops working immediately.
- [toloka project revoke-tokens](https://docs.platform.toloka.ai/docs/cli/reference/project/revoke-tokens.md): Revoke every live project-scoped token of the project; they stop working immediately.
- [toloka project create](https://docs.platform.toloka.ai/docs/cli/reference/project/create.md): Create a live project and select it.
- [toloka audience](https://docs.platform.toloka.ai/docs/cli/reference/audience.md): Inspect the expert audience catalog used by node audience filters.
- [toloka audience specialization](https://docs.platform.toloka.ai/docs/cli/reference/audience/specialization.md): Expert specializations (platform qualifications).
- [toloka audience specialization list](https://docs.platform.toloka.ai/docs/cli/reference/audience/specialization/list.md): List specializations with their id, name, and expert count.
- [toloka generation-model](https://docs.platform.toloka.ai/docs/cli/reference/generation-model.md): Inspect the models synthetic generation nodes can use.
- [toloka generation-model list](https://docs.platform.toloka.ai/docs/cli/reference/generation-model/list.md): List generation models with reasoning efforts and maxCostPerAttempt bounds.
- [toloka init](https://docs.platform.toloka.ai/docs/cli/reference/init.md): Initialize this project directory and install the agent bootstrap.
- [toloka import](https://docs.platform.toloka.ai/docs/cli/reference/import.md): Generate TypeScript source from a live project.
- [toloka plan](https://docs.platform.toloka.ai/docs/cli/reference/plan.md): Show what apply would change.
- [toloka validate](https://docs.platform.toloka.ai/docs/cli/reference/validate.md): Run platform spec validation and report matching checks.
- [toloka check](https://docs.platform.toloka.ai/docs/cli/reference/check.md): Run the complete local source, type, and test gate.
- [toloka apply](https://docs.platform.toloka.ai/docs/cli/reference/apply.md): Check source, then make the platform match it.
- [toloka secret](https://docs.platform.toloka.ai/docs/cli/reference/secret.md): Manage project secrets.
- [toloka secret set](https://docs.platform.toloka.ai/docs/cli/reference/secret/set.md): Set a secret value without writing it to disk.
- [toloka expert-list](https://docs.platform.toloka.ai/docs/cli/reference/expert-list.md): Manage the project's expert lists, which nodes with audienceMode service take experts from.
- [toloka expert-list list](https://docs.platform.toloka.ai/docs/cli/reference/expert-list/list.md): List the project's expert lists.
- [toloka expert-list create](https://docs.platform.toloka.ai/docs/cli/reference/expert-list/create.md): Create an empty expert list.
- [toloka expert-list show](https://docs.platform.toloka.ai/docs/cli/reference/expert-list/show.md): Show a list and how many of its experts are available.
- [toloka expert-list rename](https://docs.platform.toloka.ai/docs/cli/reference/expert-list/rename.md): Rename a list.
- [toloka expert-list hide](https://docs.platform.toloka.ai/docs/cli/reference/expert-list/hide.md): Hide a list: its experts can't be added or removed until it is shown again; nodes that use it keep working.
- [toloka expert-list unhide](https://docs.platform.toloka.ai/docs/cli/reference/expert-list/unhide.md): Show a hidden list again.
- [toloka expert-list experts](https://docs.platform.toloka.ai/docs/cli/reference/expert-list/experts.md): List a page of the list's experts with their status.
- [toloka expert-list add](https://docs.platform.toloka.ai/docs/cli/reference/expert-list/add.md): Add experts to a list, from arguments or a file.
- [toloka expert-list remove](https://docs.platform.toloka.ai/docs/cli/reference/expert-list/remove.md): Remove experts from a list.
- [toloka expert-ban-list](https://docs.platform.toloka.ai/docs/cli/reference/expert-ban-list.md): Manage the project's expert ban list: banned experts can't work on the project.
- [toloka expert-ban-list show](https://docs.platform.toloka.ai/docs/cli/reference/expert-ban-list/show.md): Show how many experts are banned and in which status.
- [toloka expert-ban-list experts](https://docs.platform.toloka.ai/docs/cli/reference/expert-ban-list/experts.md): List a page of the banned experts with their status.
- [toloka expert-ban-list add](https://docs.platform.toloka.ai/docs/cli/reference/expert-ban-list/add.md): Ban experts: they can't be invited and are removed from the work they're in.
- [toloka expert-ban-list remove](https://docs.platform.toloka.ai/docs/cli/reference/expert-ban-list/remove.md): Lift the bans of experts.
- [toloka base-model](https://docs.platform.toloka.ai/docs/cli/reference/base-model.md): Inspect the global base model catalog.
- [toloka base-model list](https://docs.platform.toloka.ai/docs/cli/reference/base-model/list.md): List base models.
- [toloka base-model get](https://docs.platform.toloka.ai/docs/cli/reference/base-model/get.md): Show one base model.
- [toloka model](https://docs.platform.toloka.ai/docs/cli/reference/model.md): Manage project Models and immutable versions.
- [toloka model list](https://docs.platform.toloka.ai/docs/cli/reference/model/list.md): List up to 100 Models owned by the selected project (first page only).
- [toloka model get](https://docs.platform.toloka.ai/docs/cli/reference/model/get.md): Show one project Model.
- [toloka model versions](https://docs.platform.toloka.ai/docs/cli/reference/model/versions.md): List immutable versions of a model.
- [toloka model create](https://docs.platform.toloka.ai/docs/cli/reference/model/create.md): Create project Model metadata without a Model Version.
- [toloka model archive](https://docs.platform.toloka.ai/docs/cli/reference/model/archive.md): Archive a project Model.
- [toloka model version](https://docs.platform.toloka.ai/docs/cli/reference/model/version.md): Manage one immutable Model Version.
- [toloka model version download](https://docs.platform.toloka.ai/docs/cli/reference/model/version/download.md): Download and verify Model Version content.
- [toloka training](https://docs.platform.toloka.ai/docs/cli/reference/training.md): Manage project Training drafts.
- [toloka training list](https://docs.platform.toloka.ai/docs/cli/reference/training/list.md): List up to 100 active Trainings owned by the selected project (first page only).
- [toloka training get](https://docs.platform.toloka.ai/docs/cli/reference/training/get.md): Show one Training draft.
- [toloka training create](https://docs.platform.toloka.ai/docs/cli/reference/training/create.md): Create a Training draft.
- [toloka training update](https://docs.platform.toloka.ai/docs/cli/reference/training/update.md): Update a Training draft. Omitted hyperparameters keep their current values.
- [toloka training cost](https://docs.platform.toloka.ai/docs/cli/reference/training/cost.md): Validate Training inputs and calculate the maximum charge.
- [toloka training run](https://docs.platform.toloka.ai/docs/cli/reference/training/run.md): Calculate a ready Cost Estimate and create one queued Training Run.
- [toloka training runs](https://docs.platform.toloka.ai/docs/cli/reference/training/runs.md): List up to 100 Training Runs (first page only).
- [toloka training status](https://docs.platform.toloka.ai/docs/cli/reference/training/status.md): Show one Training Run.
- [toloka training resume](https://docs.platform.toloka.ai/docs/cli/reference/training/resume.md): Resume a Training Run blocked on funds or payment.
- [toloka training cancel](https://docs.platform.toloka.ai/docs/cli/reference/training/cancel.md): Cancel a Training Run before training completes.
- [toloka training metrics](https://docs.platform.toloka.ai/docs/cli/reference/training/metrics.md): Read Training Run metrics.
- [toloka training logs](https://docs.platform.toloka.ai/docs/cli/reference/training/logs.md): Read sanitized Training Run logs.
- [toloka batch-inference](https://docs.platform.toloka.ai/docs/cli/reference/batch-inference.md): Answer dataset rows with a base model or Model Version.
- [toloka batch-inference run](https://docs.platform.toloka.ai/docs/cli/reference/batch-inference/run.md): Quote a Batch Inference Run, then confirm it to start.
- [toloka batch-inference list](https://docs.platform.toloka.ai/docs/cli/reference/batch-inference/list.md): List up to 100 of the selected project's Batch Inference Runs (first page only).
- [toloka batch-inference status](https://docs.platform.toloka.ai/docs/cli/reference/batch-inference/status.md): Show one Batch Inference Run.
- [toloka batch-inference confirm](https://docs.platform.toloka.ai/docs/cli/reference/batch-inference/confirm.md): Accept the quoted maximum charge and start the run.
- [toloka batch-inference resume](https://docs.platform.toloka.ai/docs/cli/reference/batch-inference/resume.md): Resume a run blocked on funds or payment.
- [toloka batch-inference cancel](https://docs.platform.toloka.ai/docs/cli/reference/batch-inference/cancel.md): Cancel a run; answers already generated are kept and the GPU time used is charged.
- [toloka dataset](https://docs.platform.toloka.ai/docs/cli/reference/dataset.md): Inspect and move dataset catalog and rows.
- [toloka dataset list](https://docs.platform.toloka.ai/docs/cli/reference/dataset/list.md): List live datasets in the selected project.
- [toloka dataset create](https://docs.platform.toloka.ai/docs/cli/reference/dataset/create.md): Create an empty dataset in the selected project.
- [toloka dataset fields](https://docs.platform.toloka.ai/docs/cli/reference/dataset/fields.md): Print field declarations inferred from uploaded rows, or add dataset fields from them.
- [toloka dataset fields declarations](https://docs.platform.toloka.ai/docs/cli/reference/dataset/fields/declarations.md): Print field declarations inferred from uploaded rows.
- [toloka dataset fields add](https://docs.platform.toloka.ai/docs/cli/reference/dataset/fields/add.md): Add dataset fields from fields of the uploaded rows, so filters, `--balance-by` and split columns can use their values; uploads after this fill them too.
- [toloka dataset load](https://docs.platform.toloka.ai/docs/cli/reference/dataset/load.md): Download live dataset rows into a JSONL file, one object of the chosen fields per row.
- [toloka dataset upload](https://docs.platform.toloka.ai/docs/cli/reference/dataset/upload.md): Append rows to a dataset.
- [toloka dataset slice](https://docs.platform.toloka.ai/docs/cli/reference/dataset/slice.md): Freeze the rows of a dataset, with chosen columns read as they were then.
- [toloka dataset slice create](https://docs.platform.toloka.ai/docs/cli/reference/dataset/slice/create.md): Freeze the rows that match a filter, or every row, so the same slice always selects the same rows.
- [toloka dataset slice list](https://docs.platform.toloka.ai/docs/cli/reference/dataset/slice/list.md): List up to 100 slices of a dataset, newest first (first page only).
- [toloka dataset slice show](https://docs.platform.toloka.ai/docs/cli/reference/dataset/slice/show.md): Show the status, filter, frozen columns and row count of a slice.
- [toloka dataset split](https://docs.platform.toloka.ai/docs/cli/reference/dataset/split.md): Put every row without a part in one, sampled by share; rows that already have a part keep it, so test rows stay test as the dataset grows.
- [toloka dataset columns](https://docs.platform.toloka.ai/docs/cli/reference/dataset/columns.md): List the columns of a dataset, with the parts and roles of split columns.
- [toloka pipeline](https://docs.platform.toloka.ai/docs/cli/reference/pipeline.md): Operate on a live pipeline in the selected project.
- [toloka pipeline list](https://docs.platform.toloka.ai/docs/cli/reference/pipeline/list.md): List live pipelines in the selected project.
- [toloka pipeline run](https://docs.platform.toloka.ai/docs/cli/reference/pipeline/run.md): Start a run on an applied pipeline.
- [toloka pipeline lint](https://docs.platform.toloka.ai/docs/cli/reference/pipeline/lint.md): Run the pipeline's lints that have no current result.
- [toloka pipeline status](https://docs.platform.toloka.ai/docs/cli/reference/pipeline/status.md): Report the latest run's state and item counts.
- [toloka pipeline runs](https://docs.platform.toloka.ai/docs/cli/reference/pipeline/runs.md): List recent runs.
- [toloka pipeline cost](https://docs.platform.toloka.ai/docs/cli/reference/pipeline/cost.md): Forecast what a run of this pipeline would bill.
- [toloka pipeline cancel](https://docs.platform.toloka.ai/docs/cli/reference/pipeline/cancel.md): Request cancellation of a run.
- [toloka pipeline results](https://docs.platform.toloka.ai/docs/cli/reference/pipeline/results.md): Report a run's items.
- [toloka status](https://docs.platform.toloka.ai/docs/cli/reference/status.md): Report project health: live counts and pending source changes.
- [toloka report](https://docs.platform.toloka.ai/docs/cli/reference/report.md): Preview and send an issue report to Toloka.
- [toloka skill](https://docs.platform.toloka.ai/docs/cli/reference/skill.md): Discover bundled agent skills and install them into .agents/skills.
- [toloka skill list](https://docs.platform.toloka.ai/docs/cli/reference/skill/list.md): List bundled skills as not-installed, current, or stale.
- [toloka skill search](https://docs.platform.toloka.ai/docs/cli/reference/skill/search.md): Find bundled skills for a task.
- [toloka skill install](https://docs.platform.toloka.ai/docs/cli/reference/skill/install.md): Copy one bundled skill into .agents/skills/<skillId>.
- [toloka ui](https://docs.platform.toloka.ai/docs/cli/reference/ui.md): Inspect components available for labeling interfaces.
- [toloka ui components](https://docs.platform.toloka.ai/docs/cli/reference/ui/components.md): List Task UI components, or print the props, value shape, and example of the named ones.
- [HTTP API](https://docs.platform.toloka.ai/docs/api-reference/index.md): Authenticate once and use the beta HTTP API for shared projects and datasets, Data Labeling pipelines, and Train resources.
- [Get the current user](https://docs.platform.toloka.ai/docs/api-reference/account/get-current-user.md): Return the API key owner's user id, email, and teams with their write access. Pass a writable team's id as `teamId` when creating a project.
- [List projects](https://docs.platform.toloka.ai/docs/api-reference/projects/list-projects.md): Return a page of accessible projects across your teams, with pagination metadata.
- [Create a project](https://docs.platform.toloka.ai/docs/api-reference/projects/create-project.md): Send a `teamId` and optional name. Returns the active project. Use a team where your API key has write access.
- [Get a project](https://docs.platform.toloka.ai/docs/api-reference/projects/get-project-by-id.md): Return the project name, owning team, status, and creation time. Returns 404 if the project is unavailable to you.
- [Update a project](https://docs.platform.toloka.ai/docs/api-reference/projects/update-project.md): Send the new project name. Returns the updated project.
- [Issue a project-scoped token](https://docs.platform.toloka.ai/docs/api-reference/projects/create-project-token.md): Send a permission preset or explicit permissions to create a token limited to this project and any listed `projectIds`. Send the returned token in the `X-Scoped-Token` header. Tokens last at most six hours and cannot issue other tokens.
- [List live project-scoped tokens](https://docs.platform.toloka.ai/docs/api-reference/projects/list-project-tokens.md): Return a page of tokens that have not expired or been revoked, newest first. Only IDs and metadata are returned. Requires an API key or signed-in user.
- [Revoke all project tokens](https://docs.platform.toloka.ai/docs/api-reference/projects/revoke-project-tokens.md): Revoke every unexpired token for this project and return the count. Tokens covering other projects are revoked in full. Requires an API key or signed-in user.
- [Revoke a project-scoped token](https://docs.platform.toloka.ai/docs/api-reference/projects/revoke-project-token.md): Revoke the specified token and return 204. It stops working immediately. Requires an API key or signed-in user.
- [Revoke the current token](https://docs.platform.toloka.ai/docs/api-reference/projects/revoke-current-token.md): Revoke the token sent in `X-Scoped-Token` and return 204. Returns 403 if the request uses another authentication method.
- [List project datasets](https://docs.platform.toloka.ai/docs/api-reference/datasets/list-project-datasets.md): Return a page of datasets in the project, with pagination metadata.
- [Create a dataset](https://docs.platform.toloka.ai/docs/api-reference/datasets/create-dataset.md): Send an optional name and profile. A profile creates typed fields and mappings that validate file uploads. Without a profile, uploaded rows are stored in the "uploaded" field. Returns the dataset.
- [Get a dataset](https://docs.platform.toloka.ai/docs/api-reference/datasets/get-dataset-by-id.md): Return the dataset name, project, profile, and creation time. Use the fields or items endpoints to inspect its contents.
- [List dataset fields](https://docs.platform.toloka.ai/docs/api-reference/datasets/list-dataset-fields.md): Return field definitions, split status, and active write operations. Omit `names` for all fields, or send a comma-separated list of field names.
- [Create dataset fields](https://docs.platform.toloka.ai/docs/api-reference/datasets/create-dataset-field.md): Send a field name and JSON Schema, or a `fields` array for several fields. Returns the created field or fields. Names and schemas cannot be changed through this API.
- [List inferred dataset fields](https://docs.platform.toloka.ai/docs/api-reference/datasets/get-dataset-inferred-fields.md): Returns nested field names, paths, and JSON Schemas inferred from the first file upload that passes validation. Later uploads are checked against that schema and do not refresh it. The `fields` array is empty until this inference completes.
- [Add rows to a dataset](https://docs.platform.toloka.ai/docs/api-reference/datasets/add-dataset-items.md): Send up to 10,000 row objects in a body no larger than 2 MiB. Use `fields` to map row keys to dataset fields; omitted or empty `fields` stores each complete row in "uploaded". Returns the number created. Use a file upload for larger requests.
- [List dataset items](https://docs.platform.toloka.ai/docs/api-reference/datasets/list-dataset-items.md): Return a page of item IDs, indexes, and creation times, with at most 100 items per page. Field values are available through dataset exports.
- [Mark a deterministic sample of rows](https://docs.platform.toloka.ai/docs/api-reference/datasets/sample-dataset-items.md): Send `markField`, `markValue`, and exactly one of `count` or `fraction`, with an optional filter. Marks a repeatable subset of rows in the background and returns a sampling ID. Increasing the size extends the existing sample.
- [Get dataset sampling status](https://docs.platform.toloka.ai/docs/api-reference/datasets/get-dataset-sample.md): Return the number of marked rows and the target. Poll until `status` is `succeeded`; the final count can be below the target if fewer rows match.
- [Create a dataset slice](https://docs.platform.toloka.ai/docs/api-reference/datasets/create-dataset-slice.md): Freeze the selected rows and fields at one cutoff. Use `fromSlice` to reuse an existing slice's filter at a new cutoff. Split fields are always frozen. Returns a slice with status `creating`; poll until `ready` or `failed`. Uploads and blocking field operations prevent creation.
- [List dataset slices](https://docs.platform.toloka.ai/docs/api-reference/datasets/list-dataset-slices.md): Return a page of slices, newest first, with their status and row counts.
- [Get a dataset slice](https://docs.platform.toloka.ai/docs/api-reference/datasets/get-dataset-slice.md): Poll a slice while it is being created; read its row count once it is ready.
- [Split a dataset into parts](https://docs.platform.toloka.ai/docs/api-reference/datasets/split-dataset.md): Send an optional split field name in `column`, parts, and row selection. Assigns unassigned rows to named parts while keeping existing assignments. Returns the split field; poll its split status until `ready` or `failed`. To train on the result, create a slice that freezes this field.
- [Create fields from uploaded rows](https://docs.platform.toloka.ai/docs/api-reference/datasets/create-dataset-fields-from-uploaded-rows.md): Send top-level names from the inferred fields list. Creates dataset fields with inferred JSON Schemas and copies their values from uploaded rows. Future file uploads populate them too. Splits and slices that read these fields are blocked until copying finishes.
- [Start a dataset file upload](https://docs.platform.toloka.ai/docs/api-reference/dataset-uploads/create-dataset-upload.md): Send `filename`, `declaredSizeBytes`, and an optional format and field mapping. Files may be up to 1 GiB. Returns an upload and `uploadUrl`. PUT the file bytes to that URL with the required `x-ms-blob-type: BlockBlob` header. After the PUT succeeds, finalize the upload. The file must not exceed its…
- [Finalize a dataset file upload](https://docs.platform.toloka.ai/docs/api-reference/dataset-uploads/finalize-dataset-upload.md): After uploading the file to `uploadUrl`, call this endpoint to begin importing its rows. Returns the upload status to poll. A missing file or an already finalized, failed, or canceled upload returns 400. Exceeding the declared size or 1 GiB limit returns 409 and cancels the upload.
- [Cancel a dataset upload](https://docs.platform.toloka.ai/docs/api-reference/dataset-uploads/cancel-dataset-upload.md): Cancel an upload before finalization to allow another upload to the dataset. Returns 204, or 409 if the upload has already been finalized.
- [Get a dataset upload](https://docs.platform.toloka.ai/docs/api-reference/dataset-uploads/get-dataset-upload.md): Return the upload status, imported row count, and error if it failed. After finalizing, poll until `status` is `succeeded` or `failed`.
- [Start a dataset export](https://docs.platform.toloka.ai/docs/api-reference/dataset-exports/create-dataset-export.md): Send field names or IDs in `fields`. Returns an export to poll for a download link. The file is a JSON array of dataset values; it does not contain pipeline run outputs.
- [Get a dataset export](https://docs.platform.toloka.ai/docs/api-reference/dataset-exports/get-dataset-export.md): Returns export status and any error. Poll until `status` is `succeeded` or `failed`. Successful file exports include `result.downloadUrl`. Exports prepared for training cost estimates validate the inputs without creating a file, so their download URL is null.
- [List project pipelines](https://docs.platform.toloka.ai/docs/api-reference/pipelines/list-project-pipelines.md): Return a page of pipelines in the project, with their versions and pagination metadata.
- [Get a pipeline](https://docs.platform.toloka.ai/docs/api-reference/pipelines/get-pipeline-by-id.md): Return the pipeline name, project, version, and timestamps. Use the graph endpoint to read its configuration.
- [Update a pipeline](https://docs.platform.toloka.ai/docs/api-reference/pipelines/update-pipeline.md): Send the new pipeline name. Returns the updated pipeline.
- [Import a pipeline into a project](https://docs.platform.toloka.ai/docs/api-reference/pipelines/import-pipeline.md): Send an exported pipeline document containing `name` and `graph`, up to 2 MiB. Returns the new pipeline. Returns 400 for an invalid graph or a reference to a secret missing from the project.
- [Attach a dataset to a pipeline](https://docs.platform.toloka.ai/docs/api-reference/pipelines/attach-pipeline-dataset.md): Send `datasetId` to set the Start node's source dataset. Optionally set `itemLimit` and `itemFilter`. Returns the updated pipeline; an active run prevents changes.
- [Export a pipeline](https://docs.platform.toloka.ai/docs/api-reference/pipelines/export-pipeline.md): Return a document containing the pipeline name and graph for import into another project. The attached dataset is omitted.
- [Copy a pipeline](https://docs.platform.toloka.ai/docs/api-reference/pipelines/copy-pipeline.md): Send an optional name to copy the pipeline, including its configuration and dataset binding, within its project. Returns the new pipeline's metadata. Use `GET /api/v2-beta/pipelines/{pipelineId}/graph` with the new pipeline ID to read the copied configuration.
- [Estimate pipeline cost](https://docs.platform.toloka.ai/docs/api-reference/pipelines/estimate-pipeline-cost.md): Return expected and maximum costs for the current graph and attached dataset. Optional `itemLimit` and `itemFilter` override the Start node settings for this estimate. No run is started.
- [Get the pipeline graph](https://docs.platform.toloka.ai/docs/api-reference/pipeline-graph/get-pipeline-graph-snapshot.md): Return the current graph, attached dataset, item filter, and version. Invalid stored field schemas appear in `invalidFieldSchemas` so you can repair them by replacing the graph.
- [Replace the pipeline graph](https://docs.platform.toloka.ai/docs/api-reference/pipeline-graph/replace-pipeline-graph.md): Send the complete graph, up to 2 MiB, and optionally `expectedVersion`. Returns the updated pipeline metadata. Invalid configuration or missing secrets returns 400; failed checks and lint findings do not block this write. An attached dataset cannot be removed. A stale version, concurrent edit, or ac…
- [Validate a pipeline graph](https://docs.platform.toloka.ai/docs/api-reference/pipeline-graph/validate-pipeline-graph.md): Send a graph to receive configuration findings and matching check results without saving it. Pending or running lints produce `running_node_check` warnings. A lint with no matching run produces the same warning during the first 60 seconds after graph activation, then `not_ran_node_check` afterward.…
- [Update a pipeline node](https://docs.platform.toloka.ai/docs/api-reference/pipeline-graph/patch-pipeline-node.md): Send top-level settings for one node. Nested settings are replaced as a whole. The platform validates and saves the updated graph, then returns the pipeline metadata. Use `expectedVersion` to detect conflicting edits.
- [Start a pipeline run](https://docs.platform.toloka.ai/docs/api-reference/runs/create-pipeline-run.md): Start processing the attached dataset, up to 10,000 items. Returns the run to poll. Configuration errors, failed checks, and unfinished lints on nodes from Start to End block deployment with 400. Missing lints and exam-dataset lints with findings are started when deployment is rejected; retry after…
- [List pipeline runs](https://docs.platform.toloka.ai/docs/api-reference/runs/list-pipeline-runs.md): Return a page of runs, newest first. Use each run ID to get its current status.
- [Get a pipeline run](https://docs.platform.toloka.ai/docs/api-reference/runs/get-pipeline-run.md): Return run status, item counts, and any error. Poll until `status` is `succeeded` or `failed`. A canceled run reports `failed` with `Run canceled`.
- [Get the cost of a run](https://docs.platform.toloka.ai/docs/api-reference/runs/get-pipeline-run-cost.md): Returns `total`, the cumulative run cost in US dollars as a decimal string, and `computedAt`, the ISO 8601 time when this cost query started. Poll while the run is in progress; the total never decreases.
- [Cancel a pipeline run](https://docs.platform.toloka.ai/docs/api-reference/runs/cancel-pipeline-run.md): Stop processing remaining items and return the current run state. Completed items are retained. Poll until the run reports `failed` with `Run canceled`. `pendingHumanQa` defaults to `wait_for_review`; reviews can finish after cancellation. After 14 days on the node, eligible waiting items are accept…
- [List run items](https://docs.platform.toloka.ai/docs/api-reference/runs/list-pipeline-run-items.md): Return a page of run items with status and references to input, output, and partial output field revisions.
- [Get a run item](https://docs.platform.toloka.ai/docs/api-reference/runs/get-pipeline-run-item.md): Return the item status and references to input, output, and partial output field revisions. Returns 404 if the item does not belong to this run.
- [List pipeline checks](https://docs.platform.toloka.ai/docs/api-reference/checks/list-pipeline-checks.md): Return the latest checks matching the current pipeline graph, with their status and results.
- [Start a pipeline check](https://docs.platform.toloka.ai/docs/api-reference/checks/create-pipeline-check.md): Send `nodeId` and `checkId` to check the current graph. Returns a check to poll. Returns 400 if the check is unavailable or cannot run for this graph.
- [Run pipeline lints](https://docs.platform.toloka.ai/docs/api-reference/checks/run-pipeline-lints.md): Start missing, failed, or timed-out lints for the current graph and rerun dataset-content lints. Running lints are retained. Returns checks, rejected starts, and `nodesInUse`. Poll the checks for nodes from Start to End before starting a pipeline run.
- [Get a pipeline check](https://docs.platform.toloka.ai/docs/api-reference/checks/get-pipeline-check.md): Return check status, output, and report. Poll until `status` is `succeeded` or `failed`.
- [List project secret names](https://docs.platform.toloka.ai/docs/api-reference/secrets/list-project-secrets.md): Return secret names and creation and update times. Secret values are omitted.
- [Set a project secret](https://docs.platform.toloka.ai/docs/api-reference/secrets/set-project-secret.md): Send `value` to create or replace the named secret. Returns its name and timestamps without the secret value.
- [List project files](https://docs.platform.toloka.ai/docs/api-reference/files/list-project-files.md): Return the project files with upload status, size, and download links.
- [Start a project file upload](https://docs.platform.toloka.ai/docs/api-reference/files/create-project-file-upload.md): Send `filename` and `declaredSizeBytes`, up to 100 MiB. Returns file metadata and `uploadUrl`. PUT the file bytes to that URL with the required `x-ms-blob-type: BlockBlob` header. After the PUT succeeds, finalize the upload.
- [Finalize a project file upload](https://docs.platform.toloka.ai/docs/api-reference/files/finalize-project-file-upload.md): After uploading bytes to `uploadUrl`, call this endpoint to verify the file. Returns updated file metadata. A missing file or size above 100 MiB returns 400; a failed upload or a file that changes during verification returns 409.
- [Delete a project file](https://docs.platform.toloka.ai/docs/api-reference/files/delete-project-file.md): Remove the file from the project. Returns 204, or 404 if the project or file is unavailable to you.
- [List synthetic generation models](https://docs.platform.toloka.ai/docs/api-reference/generation-models/list-generation-models.md): Return the available models, supported reasoning efforts, and cost limits for synthetic generation. Use `id` as `generationModelId`. Set `reasoningEffort` only when `defaultReasoningEffort` is present. `maxCostPerAttempt` must be between `min` and `maxEstimate`; values below `typical` produce a warn…
- [List expert specializations](https://docs.platform.toloka.ai/docs/api-reference/audience/list-audience-specializations.md): Return a page of expert specializations and eligible expert counts, sorted by name and ID. The default and maximum page size is 1000. Use each result's `id` as `qualificationId` and `name` as `value` in `audience.specializations.values[]`.
- [List expert lists](https://docs.platform.toloka.ai/docs/api-reference/expert-lists/list-expert-lists.md): Return the expert lists of a project, newest first. Available only to teams with expert lists enabled.
- [Create an expert list](https://docs.platform.toloka.ai/docs/api-reference/expert-lists/create-expert-list.md): Create an empty expert list in a project. Add experts with `addExpertListExperts`.
- [Get an expert list](https://docs.platform.toloka.ai/docs/api-reference/expert-lists/get-expert-list.md): Return an expert list with how many of its experts are in each status.
- [Rename, hide or show an expert list](https://docs.platform.toloka.ai/docs/api-reference/expert-lists/update-expert-list.md)
- [List the experts of an expert list](https://docs.platform.toloka.ai/docs/api-reference/expert-lists/list-expert-list-experts.md): Return a page of the experts in a list with their status, sorted by expert ID.
- [Add experts to an expert list](https://docs.platform.toloka.ai/docs/api-reference/expert-lists/add-expert-list-experts.md)
- [Remove experts from an expert list](https://docs.platform.toloka.ai/docs/api-reference/expert-lists/remove-expert-list-experts.md)
- [Get the expert ban list of a project](https://docs.platform.toloka.ai/docs/api-reference/expert-ban-list/get-expert-ban-list.md): Return how many experts are in the project's own ban list and how many are in each status. Banned experts cannot work on the project. Team-wide bans also apply but are managed in the platform UI and not counted here. Available only to teams with expert bans enabled.
- [List the experts of the expert ban list](https://docs.platform.toloka.ai/docs/api-reference/expert-ban-list/list-expert-ban-list-experts.md): Return a page of the experts in the project's own ban list with their status, sorted by expert ID. Team-wide bans are not listed.
- [Ban experts in a project](https://docs.platform.toloka.ai/docs/api-reference/expert-ban-list/add-expert-ban-list-experts.md): Ban experts so they cannot be invited to the project and are removed from the work they are already in. The request is rejected when any of the experts is already banned.
- [Unban experts in a project](https://docs.platform.toloka.ai/docs/api-reference/expert-ban-list/remove-expert-ban-list-experts.md)
- [List base models](https://docs.platform.toloka.ai/docs/api-reference/base-models/list-base-models.md): Return the base model catalog, including model names and fixed revisions.
- [Get a base model](https://docs.platform.toloka.ai/docs/api-reference/base-models/get-base-model.md): Return the name and fixed revision of a catalog model. Returns 404 if the ID is unknown.
- [Create a training](https://docs.platform.toloka.ai/docs/api-reference/trainings/create-project-training.md): Send a name and configuration to create a training draft. The configuration requires a ready dataset slice that freezes a split field. Returns the draft; its output model is created when the first run starts.
- [Get a training](https://docs.platform.toloka.ai/docs/api-reference/trainings/get-training.md): Return the training metadata and saved configuration, including applied defaults and dataset selections.
- [List project trainings](https://docs.platform.toloka.ai/docs/api-reference/trainings/list-project-trainings.md): Return a page of non-archived training drafts in the project, with pagination metadata.
- [Update training metadata](https://docs.platform.toloka.ai/docs/api-reference/trainings/update-training-metadata.md): Send a new name or archive status. Returns the updated training without changing its configuration.
- [Replace a training configuration](https://docs.platform.toloka.ai/docs/api-reference/trainings/replace-training-configuration.md): Send the complete configuration, including a ready dataset slice that freezes a split field. Returns the updated training. Concurrent changes use the last write; archived trainings cannot be updated.
- [Prepare training cost estimate inputs](https://docs.platform.toloka.ai/docs/api-reference/trainings/prepare-training-cost-estimate-inputs.md): Send an `Idempotency-Key` header and an empty JSON body to freeze training and validation inputs for the current configuration. Poll each non-null ID in `exports` with `GET /api/v2-beta/exports/{exportId}` until its `status` is `succeeded` or `failed`. A failed export blocks the estimate with an `ex…
- [Estimate training cost](https://docs.platform.toloka.ai/docs/api-reference/trainings/estimate-training-cost.md): Send the prepared `inputSetId`. Returns either a maximum charge and fingerprint for starting a run, or issues that block training. Returns 409 while inputs are still being prepared or after they expire.
- [Create a training run](https://docs.platform.toloka.ai/docs/api-reference/trainings/create-training-run.md): Send `inputSetId`, `estimateFingerprint`, `versionName`, and an `Idempotency-Key` header. The platform checks the estimate again and returns the training run. Returns 409 if the estimate changed, the input set is in use, the version name exists, or the key was used for different input.
- [Get a training run](https://docs.platform.toloka.ai/docs/api-reference/trainings/get-training-run.md): Return run status, progress, billing details, and input selections. Poll until `completed`, `failed`, or `canceled`. If `blocked`, inspect `blockedReason` and `message`.
- [List training runs](https://docs.platform.toloka.ai/docs/api-reference/trainings/list-training-runs.md): Return a page of runs for this training, with their configuration, status, and pagination metadata.
- [Read training run logs](https://docs.platform.toloka.ai/docs/api-reference/trainings/get-training-run-logs.md): Return up to 100 log entries. Pass `nextCursor` as `cursor` to continue reading; `complete` indicates whether the log is finished.
- [Get training run metrics](https://docs.platform.toloka.ai/docs/api-reference/trainings/get-training-run-metrics.md): Return the latest recorded value of each metric for the run, including its step and timestamp.
- [Get training metric history](https://docs.platform.toloka.ai/docs/api-reference/trainings/get-training-run-metric-history.md): Use a value from `metrics[].key` returned by `GET /api/v2-beta/training-runs/{trainingRunId}/metrics` as `metricKey`. Set `resolution` to choose the number of sampled history points; the default is 200. Returns 404 if the run or metric is unavailable.
- [Cancel a training run](https://docs.platform.toloka.ai/docs/api-reference/trainings/cancel-training-run.md): Request cancellation and return the current run state. Running work may report `stopping` before `canceled`. Returns 409 if output processing or payment prevents cancellation. Finished runs are returned unchanged.
- [Resume a training run](https://docs.platform.toloka.ai/docs/api-reference/trainings/resume-training-run.md): Retry a run blocked on funds or payment after adding funds. Returns its updated state. Returns 409 if the run is not resumable or its payment window has expired.
- [List project models](https://docs.platform.toloka.ai/docs/api-reference/models/list-project-models.md): Return a page of non-archived models in the project, with pagination metadata.
- [Create a project model](https://docs.platform.toloka.ai/docs/api-reference/models/create-project-model.md): Send a name and `baseModelId` to create a model in the project. Returns its metadata; model versions are created separately.
- [List model versions](https://docs.platform.toloka.ai/docs/api-reference/models/list-model-versions.md): Return a page of immutable versions belonging to the specified model, with pagination metadata.
- [Get a model version](https://docs.platform.toloka.ai/docs/api-reference/models/get-model-version.md): Return the immutable version metadata, file manifest, and source training run.
- [Get model version download links](https://docs.platform.toloka.ai/docs/api-reference/models/get-model-version-content.md): Return the version metadata and a temporary download link for each file in its manifest.
- [Archive a project model](https://docs.platform.toloka.ai/docs/api-reference/models/archive-project-model.md): Archive the model so it no longer appears in the project model list. Returns the archived model metadata.
- [Get a project model](https://docs.platform.toloka.ai/docs/api-reference/models/get-project-model.md): Return model metadata, including its base model and archive status.
- [Create a batch inference run](https://docs.platform.toloka.ai/docs/api-reference/batch-inference/create-batch-inference-run.md): Send the dataset, model, sampling settings, and output destination, with an `Idempotency-Key` header. Returns a run that prepares a quote for frozen input rows. Confirm the quote before funds are reserved or inference starts.
- [List batch inference runs](https://docs.platform.toloka.ai/docs/api-reference/batch-inference/list-batch-inference-runs.md): Return a page of batch inference runs in the project, newest first, with pagination metadata.
- [Get a batch inference run](https://docs.platform.toloka.ai/docs/api-reference/batch-inference/get-batch-inference-run.md): Return status, progress, quote, charge, and output location. Poll until the quote needs confirmation or the run reaches `completed`, `failed`, or `canceled`. If `blocked`, inspect `blockedReason` and `message`.
- [Confirm a batch inference run](https://docs.platform.toloka.ai/docs/api-reference/batch-inference/confirm-batch-inference-run.md): Accept the prepared quote to reserve its maximum charge and start inference. Returns the updated run. Returns 409 if the quote is unavailable or expired, or the run failed or was canceled. Repeated confirmation of an already confirmed run returns its current state.
- [Resume a batch inference run](https://docs.platform.toloka.ai/docs/api-reference/batch-inference/resume-batch-inference-run.md): After adding funds, retry a run blocked on funds or payment. Returns its updated state. Payment retries are allowed for seven days after the first attempt; an expired payment window returns 409.
- [Cancel a batch inference run](https://docs.platform.toloka.ai/docs/api-reference/batch-inference/cancel-batch-inference-run.md): Request cancellation and return the current run state. Active work may report `stopping` before `canceled`. Returns 409 during payment, billing review, or output writing. Finished runs are returned unchanged.
- [Troubleshoot a pipeline run](https://docs.platform.toloka.ai/docs/data/guides/troubleshoot.md): Find fixes for stuck runs, blocked deployments, unexpected charges, missing export results, and upload problems.
- [Pipeline errors and how to fix them](https://docs.platform.toloka.ai/docs/data/reference/errors.md): Find an error message by screen, identify its cause, and follow the fix.
- [Training errors](https://docs.platform.toloka.ai/docs/train/reference/errors.md): Find exact CLI, input-validation, cost-estimate, and Training Run errors with corrective actions.

## OpenAPI Specs

- [openapi](/docs/api-reference/openapi.json)


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