openai/src/resources/beta/threads/runs/runs.ts

The official TypeScript library for the OpenAI API

// File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.

import { APIResource } from '../../../../resource';
import { isRequestOptions } from '../../../../core';
import { APIPromise } from '../../../../core';
import * as Core from '../../../../core';
import { AssistantStream, RunCreateParamsBaseStream } from '../../../../lib/AssistantStream';
import { sleep } from '../../../../core';
import { RunSubmitToolOutputsParamsStream } from '../../../../lib/AssistantStream';
import * as RunsAPI from './runs';
import * as AssistantsAPI from '../../assistants';
import * as ChatAPI from '../../../chat/chat';
import * as MessagesAPI from '../messages';
import * as ThreadsAPI from '../threads';
import * as StepsAPI from './steps';
import {
  CodeInterpreterLogs,
  CodeInterpreterOutputImage,
  CodeInterpreterToolCall,
  CodeInterpreterToolCallDelta,
  FileSearchToolCall,
  FileSearchToolCallDelta,
  FunctionToolCall,
  FunctionToolCallDelta,
  MessageCreationStepDetails,
  RunStep,
  RunStepDelta,
  RunStepDeltaEvent,
  RunStepDeltaMessageDelta,
  RunStepInclude,
  RunStepsPage,
  StepListParams,
  StepRetrieveParams,
  Steps,
  ToolCall,
  ToolCallDelta,
  ToolCallDeltaObject,
  ToolCallsStepDetails,
} from './steps';
import { CursorPage, type CursorPageParams } from '../../../../pagination';
import { Stream } from '../../../../streaming';

export class Runs extends APIResource {
  steps: StepsAPI.Steps = new StepsAPI.Steps(this._client);

  /**
   * Create a run.
   */
  create(
    threadId: string,
    params: RunCreateParamsNonStreaming,
    options?: Core.RequestOptions,
  ): APIPromise<Run>;
  create(
    threadId: string,
    params: RunCreateParamsStreaming,
    options?: Core.RequestOptions,
  ): APIPromise<Stream<AssistantsAPI.AssistantStreamEvent>>;
  create(
    threadId: string,
    params: RunCreateParamsBase,
    options?: Core.RequestOptions,
  ): APIPromise<Stream<AssistantsAPI.AssistantStreamEvent> | Run>;
  create(
    threadId: string,
    params: RunCreateParams,
    options?: Core.RequestOptions,
  ): APIPromise<Run> | APIPromise<Stream<AssistantsAPI.AssistantStreamEvent>> {
    const { include, ...body } = params;
    return this._client.post(`/threads/${threadId}/runs`, {
      query: { include },
      body,
      ...options,
      headers: { 'OpenAI-Beta': 'assistants=v2', ...options?.headers },
      stream: params.stream ?? false,
    }) as APIPromise<Run> | APIPromise<Stream<AssistantsAPI.AssistantStreamEvent>>;
  }

  /**
   * Retrieves a run.
   */
  retrieve(threadId: string, runId: string, options?: Core.RequestOptions): Core.APIPromise<Run> {
    return this._client.get(`/threads/${threadId}/runs/${runId}`, {
      ...options,
      headers: { 'OpenAI-Beta': 'assistants=v2', ...options?.headers },
    });
  }

  /**
   * Modifies a run.
   */
  update(
    threadId: string,
    runId: string,
    body: RunUpdateParams,
    options?: Core.RequestOptions,
  ): Core.APIPromise<Run> {
    return this._client.post(`/threads/${threadId}/runs/${runId}`, {
      body,
      ...options,
      headers: { 'OpenAI-Beta': 'assistants=v2', ...options?.headers },
    });
  }

  /**
   * Returns a list of runs belonging to a thread.
   */
  list(
    threadId: string,
    query?: RunListParams,
    options?: Core.RequestOptions,
  ): Core.PagePromise<RunsPage, Run>;
  list(threadId: string, options?: Core.RequestOptions): Core.PagePromise<RunsPage, Run>;
  list(
    threadId: string,
    query: RunListParams | Core.RequestOptions = {},
    options?: Core.RequestOptions,
  ): Core.PagePromise<RunsPage, Run> {
    if (isRequestOptions(query)) {
      return this.list(threadId, {}, query);
    }
    return this._client.getAPIList(`/threads/${threadId}/runs`, RunsPage, {
      query,
      ...options,
      headers: { 'OpenAI-Beta': 'assistants=v2', ...options?.headers },
    });
  }

  /**
   * Cancels a run that is `in_progress`.
   */
  cancel(threadId: string, runId: string, options?: Core.RequestOptions): Core.APIPromise<Run> {
    return this._client.post(`/threads/${threadId}/runs/${runId}/cancel`, {
      ...options,
      headers: { 'OpenAI-Beta': 'assistants=v2', ...options?.headers },
    });
  }

  /**
   * A helper to create a run an poll for a terminal state. More information on Run
   * lifecycles can be found here:
   * https://platform.openai.com/docs/assistants/how-it-works/runs-and-run-steps
   */
  async createAndPoll(
    threadId: string,
    body: RunCreateParamsNonStreaming,
    options?: Core.RequestOptions & { pollIntervalMs?: number },
  ): Promise<Run> {
    const run = await this.create(threadId, body, options);
    return await this.poll(threadId, run.id, options);
  }

  /**
   * Create a Run stream
   *
   * @deprecated use `stream` instead
   */
  createAndStream(
    threadId: string,
    body: RunCreateParamsBaseStream,
    options?: Core.RequestOptions,
  ): AssistantStream {
    return AssistantStream.createAssistantStream(threadId, this._client.beta.threads.runs, body, options);
  }

  /**
   * A helper to poll a run status until it reaches a terminal state. More
   * information on Run lifecycles can be found here:
   * https://platform.openai.com/docs/assistants/how-it-works/runs-and-run-steps
   */
  async poll(
    threadId: string,
    runId: string,
    options?: Core.RequestOptions & { pollIntervalMs?: number },
  ): Promise<Run> {
    const headers: { [key: string]: string } = { ...options?.headers, 'X-Stainless-Poll-Helper': 'true' };

    if (options?.pollIntervalMs) {
      headers['X-Stainless-Custom-Poll-Interval'] = options.pollIntervalMs.toString();
    }

    while (true) {
      const { data: run, response } = await this.retrieve(threadId, runId, {
        ...options,
        headers: { ...options?.headers, ...headers },
      }).withResponse();

      switch (run.status) {
        //If we are in any sort of intermediate state we poll
        case 'queued':
        case 'in_progress':
        case 'cancelling':
          let sleepInterval = 5000;

          if (options?.pollIntervalMs) {
            sleepInterval = options.pollIntervalMs;
          } else {
            const headerInterval = response.headers.get('openai-poll-after-ms');
            if (headerInterval) {
              const headerIntervalMs = parseInt(headerInterval);
              if (!isNaN(headerIntervalMs)) {
                sleepInterval = headerIntervalMs;
              }
            }
          }
          await sleep(sleepInterval);
          break;
        //We return the run in any terminal state.
        case 'requires_action':
        case 'incomplete':
        case 'cancelled':
        case 'completed':
        case 'failed':
        case 'expired':
          return run;
      }
    }
  }

  /**
   * Create a Run stream
   */
  stream(threadId: string, body: RunCreateParamsBaseStream, options?: Core.RequestOptions): AssistantStream {
    return AssistantStream.createAssistantStream(threadId, this._client.beta.threads.runs, body, options);
  }

  /**
   * When a run has the `status: "requires_action"` and `required_action.type` is
   * `submit_tool_outputs`, this endpoint can be used to submit the outputs from the
   * tool calls once they're all completed. All outputs must be submitted in a single
   * request.
   */
  submitToolOutputs(
    threadId: string,
    runId: string,
    body: RunSubmitToolOutputsParamsNonStreaming,
    options?: Core.RequestOptions,
  ): APIPromise<Run>;
  submitToolOutputs(
    threadId: string,
    runId: string,
    body: RunSubmitToolOutputsParamsStreaming,
    options?: Core.RequestOptions,
  ): APIPromise<Stream<AssistantsAPI.AssistantStreamEvent>>;
  submitToolOutputs(
    threadId: string,
    runId: string,
    body: RunSubmitToolOutputsParamsBase,
    options?: Core.RequestOptions,
  ): APIPromise<Stream<AssistantsAPI.AssistantStreamEvent> | Run>;
  submitToolOutputs(
    threadId: string,
    runId: string,
    body: RunSubmitToolOutputsParams,
    options?: Core.RequestOptions,
  ): APIPromise<Run> | APIPromise<Stream<AssistantsAPI.AssistantStreamEvent>> {
    return this._client.post(`/threads/${threadId}/runs/${runId}/submit_tool_outputs`, {
      body,
      ...options,
      headers: { 'OpenAI-Beta': 'assistants=v2', ...options?.headers },
      stream: body.stream ?? false,
    }) as APIPromise<Run> | APIPromise<Stream<AssistantsAPI.AssistantStreamEvent>>;
  }

  /**
   * A helper to submit a tool output to a run and poll for a terminal run state.
   * More information on Run lifecycles can be found here:
   * https://platform.openai.com/docs/assistants/how-it-works/runs-and-run-steps
   */
  async submitToolOutputsAndPoll(
    threadId: string,
    runId: string,
    body: RunSubmitToolOutputsParamsNonStreaming,
    options?: Core.RequestOptions & { pollIntervalMs?: number },
  ): Promise<Run> {
    const run = await this.submitToolOutputs(threadId, runId, body, options);
    return await this.poll(threadId, run.id, options);
  }

  /**
   * Submit the tool outputs from a previous run and stream the run to a terminal
   * state. More information on Run lifecycles can be found here:
   * https://platform.openai.com/docs/assistants/how-it-works/runs-and-run-steps
   */
  submitToolOutputsStream(
    threadId: string,
    runId: string,
    body: RunSubmitToolOutputsParamsStream,
    options?: Core.RequestOptions,
  ): AssistantStream {
    return AssistantStream.createToolAssistantStream(
      threadId,
      runId,
      this._client.beta.threads.runs,
      body,
      options,
    );
  }
}

export class RunsPage extends CursorPage<Run> {}

/**
 * Tool call objects
 */
export interface RequiredActionFunctionToolCall {
  /**
   * The ID of the tool call. This ID must be referenced when you submit the tool
   * outputs in using the
   * [Submit tool outputs to run](https://platform.openai.com/docs/api-reference/runs/submitToolOutputs)
   * endpoint.
   */
  id: string;

  /**
   * The function definition.
   */
  function: RequiredActionFunctionToolCall.Function;

  /**
   * The type of tool call the output is required for. For now, this is always
   * `function`.
   */
  type: 'function';
}

export namespace RequiredActionFunctionToolCall {
  /**
   * The function definition.
   */
  export interface Function {
    /**
     * The arguments that the model expects you to pass to the function.
     */
    arguments: string;

    /**
     * The name of the function.
     */
    name: string;
  }
}

/**
 * Represents an execution run on a
 * [thread](https://platform.openai.com/docs/api-reference/threads).
 */
export interface Run {
  /**
   * The identifier, which can be referenced in API endpoints.
   */
  id: string;

  /**
   * The ID of the
   * [assistant](https://platform.openai.com/docs/api-reference/assistants) used for
   * execution of this run.
   */
  assistant_id: string;

  /**
   * The Unix timestamp (in seconds) for when the run was cancelled.
   */
  cancelled_at: number | null;

  /**
   * The Unix timestamp (in seconds) for when the run was completed.
   */
  completed_at: number | null;

  /**
   * The Unix timestamp (in seconds) for when the run was created.
   */
  created_at: number;

  /**
   * The Unix timestamp (in seconds) for when the run will expire.
   */
  expires_at: number | null;

  /**
   * The Unix timestamp (in seconds) for when the run failed.
   */
  failed_at: number | null;

  /**
   * Details on why the run is incomplete. Will be `null` if the run is not
   * incomplete.
   */
  incomplete_details: Run.IncompleteDetails | null;

  /**
   * The instructions that the
   * [assistant](https://platform.openai.com/docs/api-reference/assistants) used for
   * this run.
   */
  instructions: string;

  /**
   * The last error associated with this run. Will be `null` if there are no errors.
   */
  last_error: Run.LastError | null;

  /**
   * The maximum number of completion tokens specified to have been used over the
   * course of the run.
   */
  max_completion_tokens: number | null;

  /**
   * The maximum number of prompt tokens specified to have been used over the course
   * of the run.
   */
  max_prompt_tokens: number | null;

  /**
   * Set of 16 key-value pairs that can be attached to an object. This can be useful
   * for storing additional information about the object in a structured format. Keys
   * can be a maximum of 64 characters long and values can be a maxium of 512
   * characters long.
   */
  metadata: unknown | null;

  /**
   * The model that the
   * [assistant](https://platform.openai.com/docs/api-reference/assistants) used for
   * this run.
   */
  model: string;

  /**
   * The object type, which is always `thread.run`.
   */
  object: 'thread.run';

  /**
   * Whether to enable
   * [parallel function calling](https://platform.openai.com/docs/guides/function-calling#configuring-parallel-function-calling)
   * during tool use.
   */
  parallel_tool_calls: boolean;

  /**
   * Details on the action required to continue the run. Will be `null` if no action
   * is required.
   */
  required_action: Run.RequiredAction | null;

  /**
   * Specifies the format that the model must output. Compatible with
   * [GPT-4o](https://platform.openai.com/docs/models#gpt-4o),
   * [GPT-4 Turbo](https://platform.openai.com/docs/models#gpt-4-turbo-and-gpt-4),
   * and all GPT-3.5 Turbo models since `gpt-3.5-turbo-1106`.
   *
   * Setting to `{ "type": "json_schema", "json_schema": {...} }` enables Structured
   * Outputs which ensures the model will match your supplied JSON schema. Learn more
   * in the
   * [Structured Outputs guide](https://platform.openai.com/docs/guides/structured-outputs).
   *
   * Setting to `{ "type": "json_object" }` enables JSON mode, which ensures the
   * message the model generates is valid JSON.
   *
   * **Important:** when using JSON mode, you **must** also instruct the model to
   * produce JSON yourself via a system or user message. Without this, the model may
   * generate an unending stream of whitespace until the generation reaches the token
   * limit, resulting in a long-running and seemingly "stuck" request. Also note that
   * the message content may be partially cut off if `finish_reason="length"`, which
   * indicates the generation exceeded `max_tokens` or the conversation exceeded the
   * max context length.
   */
  response_format: ThreadsAPI.AssistantResponseFormatOption | null;

  /**
   * The Unix timestamp (in seconds) for when the run was started.
   */
  started_at: number | null;

  /**
   * The status of the run, which can be either `queued`, `in_progress`,
   * `requires_action`, `cancelling`, `cancelled`, `failed`, `completed`,
   * `incomplete`, or `expired`.
   */
  status: RunStatus;

  /**
   * The ID of the [thread](https://platform.openai.com/docs/api-reference/threads)
   * that was executed on as a part of this run.
   */
  thread_id: string;

  /**
   * Controls which (if any) tool is called by the model. `none` means the model will
   * not call any tools and instead generates a message. `auto` is the default value
   * and means the model can pick between generating a message or calling one or more
   * tools. `required` means the model must call one or more tools before responding
   * to the user. Specifying a particular tool like `{"type": "file_search"}` or
   * `{"type": "function", "function": {"name": "my_function"}}` forces the model to
   * call that tool.
   */
  tool_choice: ThreadsAPI.AssistantToolChoiceOption | null;

  /**
   * The list of tools that the
   * [assistant](https://platform.openai.com/docs/api-reference/assistants) used for
   * this run.
   */
  tools: Array<AssistantsAPI.AssistantTool>;

  /**
   * Controls for how a thread will be truncated prior to the run. Use this to
   * control the intial context window of the run.
   */
  truncation_strategy: Run.TruncationStrategy | null;

  /**
   * Usage statistics related to the run. This value will be `null` if the run is not
   * in a terminal state (i.e. `in_progress`, `queued`, etc.).
   */
  usage: Run.Usage | null;

  /**
   * The sampling temperature used for this run. If not set, defaults to 1.
   */
  temperature?: number | null;

  /**
   * The nucleus sampling value used for this run. If not set, defaults to 1.
   */
  top_p?: number | null;
}

export namespace Run {
  /**
   * Details on why the run is incomplete. Will be `null` if the run is not
   * incomplete.
   */
  export interface IncompleteDetails {
    /**
     * The reason why the run is incomplete. This will point to which specific token
     * limit was reached over the course of the run.
     */
    reason?: 'max_completion_tokens' | 'max_prompt_tokens';
  }

  /**
   * The last error associated with this run. Will be `null` if there are no errors.
   */
  export interface LastError {
    /**
     * One of `server_error`, `rate_limit_exceeded`, or `invalid_prompt`.
     */
    code: 'server_error' | 'rate_limit_exceeded' | 'invalid_prompt';

    /**
     * A human-readable description of the error.
     */
    message: string;
  }

  /**
   * Details on the action required to continue the run. Will be `null` if no action
   * is required.
   */
  export interface RequiredAction {
    /**
     * Details on the tool outputs needed for this run to continue.
     */
    submit_tool_outputs: RequiredAction.SubmitToolOutputs;

    /**
     * For now, this is always `submit_tool_outputs`.
     */
    type: 'submit_tool_outputs';
  }

  export namespace RequiredAction {
    /**
     * Details on the tool outputs needed for this run to continue.
     */
    export interface SubmitToolOutputs {
      /**
       * A list of the relevant tool calls.
       */
      tool_calls: Array<RunsAPI.RequiredActionFunctionToolCall>;
    }
  }

  /**
   * Controls for how a thread will be truncated prior to the run. Use this to
   * control the intial context window of the run.
   */
  export interface TruncationStrategy {
    /**
     * The truncation strategy to use for the thread. The default is `auto`. If set to
     * `last_messages`, the thread will be truncated to the n most recent messages in
     * the thread. When set to `auto`, messages in the middle of the thread will be
     * dropped to fit the context length of the model, `max_prompt_tokens`.
     */
    type: 'auto' | 'last_messages';

    /**
     * The number of most recent messages from the thread when constructing the context
     * for the run.
     */
    last_messages?: number | null;
  }

  /**
   * Usage statistics related to the run. This value will be `null` if the run is not
   * in a terminal state (i.e. `in_progress`, `queued`, etc.).
   */
  export interface Usage {
    /**
     * Number of completion tokens used over the course of the run.
     */
    completion_tokens: number;

    /**
     * Number of prompt tokens used over the course of the run.
     */
    prompt_tokens: number;

    /**
     * Total number of tokens used (prompt + completion).
     */
    total_tokens: number;
  }
}

/**
 * The status of the run, which can be either `queued`, `in_progress`,
 * `requires_action`, `cancelling`, `cancelled`, `failed`, `completed`,
 * `incomplete`, or `expired`.
 */
export type RunStatus =
  | 'queued'
  | 'in_progress'
  | 'requires_action'
  | 'cancelling'
  | 'cancelled'
  | 'failed'
  | 'completed'
  | 'incomplete'
  | 'expired';

export type RunCreateParams = RunCreateParamsNonStreaming | RunCreateParamsStreaming;

export interface RunCreateParamsBase {
  /**
   * Body param: The ID of the
   * [assistant](https://platform.openai.com/docs/api-reference/assistants) to use to
   * execute this run.
   */
  assistant_id: string;

  /**
   * Query param: A list of additional fields to include in the response. Currently
   * the only supported value is
   * `step_details.tool_calls[*].file_search.results[*].content` to fetch the file
   * search result content.
   *
   * See the
   * [file search tool documentation](https://platform.openai.com/docs/assistants/tools/file-search#customizing-file-search-settings)
   * for more information.
   */
  include?: Array<StepsAPI.RunStepInclude>;

  /**
   * Body param: Appends additional instructions at the end of the instructions for
   * the run. This is useful for modifying the behavior on a per-run basis without
   * overriding other instructions.
   */
  additional_instructions?: string | null;

  /**
   * Body param: Adds additional messages to the thread before creating the run.
   */
  additional_messages?: Array<RunCreateParams.AdditionalMessage> | null;

  /**
   * Body param: Overrides the
   * [instructions](https://platform.openai.com/docs/api-reference/assistants/createAssistant)
   * of the assistant. This is useful for modifying the behavior on a per-run basis.
   */
  instructions?: string | null;

  /**
   * Body param: The maximum number of completion tokens that may be used over the
   * course of the run. The run will make a best effort to use only the number of
   * completion tokens specified, across multiple turns of the run. If the run
   * exceeds the number of completion tokens specified, the run will end with status
   * `incomplete`. See `incomplete_details` for more info.
   */
  max_completion_tokens?: number | null;

  /**
   * Body param: The maximum number of prompt tokens that may be used over the course
   * of the run. The run will make a best effort to use only the number of prompt
   * tokens specified, across multiple turns of the run. If the run exceeds the
   * number of prompt tokens specified, the run will end with status `incomplete`.
   * See `incomplete_details` for more info.
   */
  max_prompt_tokens?: number | null;

  /**
   * Body param: Set of 16 key-value pairs that can be attached to an object. This
   * can be useful for storing additional information about the object in a
   * structured format. Keys can be a maximum of 64 characters long and values can be
   * a maxium of 512 characters long.
   */
  metadata?: unknown | null;

  /**
   * Body param: The ID of the
   * [Model](https://platform.openai.com/docs/api-reference/models) to be used to
   * execute this run. If a value is provided here, it will override the model
   * associated with the assistant. If not, the model associated with the assistant
   * will be used.
   */
  model?: (string & {}) | ChatAPI.ChatModel | null;

  /**
   * Body param: Whether to enable
   * [parallel function calling](https://platform.openai.com/docs/guides/function-calling#configuring-parallel-function-calling)
   * during tool use.
   */
  parallel_tool_calls?: boolean;

  /**
   * Body param: Specifies the format that the model must output. Compatible with
   * [GPT-4o](https://platform.openai.com/docs/models#gpt-4o),
   * [GPT-4 Turbo](https://platform.openai.com/docs/models#gpt-4-turbo-and-gpt-4),
   * and all GPT-3.5 Turbo models since `gpt-3.5-turbo-1106`.
   *
   * Setting to `{ "type": "json_schema", "json_schema": {...} }` enables Structured
   * Outputs which ensures the model will match your supplied JSON schema. Learn more
   * in the
   * [Structured Outputs guide](https://platform.openai.com/docs/guides/structured-outputs).
   *
   * Setting to `{ "type": "json_object" }` enables JSON mode, which ensures the
   * message the model generates is valid JSON.
   *
   * **Important:** when using JSON mode, you **must** also instruct the model to
   * produce JSON yourself via a system or user message. Without this, the model may
   * generate an unending stream of whitespace until the generation reaches the token
   * limit, resulting in a long-running and seemingly "stuck" request. Also note that
   * the message content may be partially cut off if `finish_reason="length"`, which
   * indicates the generation exceeded `max_tokens` or the conversation exceeded the
   * max context length.
   */
  response_format?: ThreadsAPI.AssistantResponseFormatOption | null;

  /**
   * Body param: If `true`, returns a stream of events that happen during the Run as
   * server-sent events, terminating when the Run enters a terminal state with a
   * `data: [DONE]` message.
   */
  stream?: boolean | null;

  /**
   * Body param: What sampling temperature to use, between 0 and 2. Higher values
   * like 0.8 will make the output more random, while lower values like 0.2 will make
   * it more focused and deterministic.
   */
  temperature?: number | null;

  /**
   * Body param: Controls which (if any) tool is called by the model. `none` means
   * the model will not call any tools and instead generates a message. `auto` is the
   * default value and means the model can pick between generating a message or
   * calling one or more tools. `required` means the model must call one or more
   * tools before responding to the user. Specifying a particular tool like
   * `{"type": "file_search"}` or
   * `{"type": "function", "function": {"name": "my_function"}}` forces the model to
   * call that tool.
   */
  tool_choice?: ThreadsAPI.AssistantToolChoiceOption | null;

  /**
   * Body param: Override the tools the assistant can use for this run. This is
   * useful for modifying the behavior on a per-run basis.
   */
  tools?: Array<AssistantsAPI.AssistantTool> | null;

  /**
   * Body param: An alternative to sampling with temperature, called nucleus
   * sampling, where the model considers the results of the tokens with top_p
   * probability mass. So 0.1 means only the tokens comprising the top 10%
   * probability mass are considered.
   *
   * We generally recommend altering this or temperature but not both.
   */
  top_p?: number | null;

  /**
   * Body param: Controls for how a thread will be truncated prior to the run. Use
   * this to control the intial context window of the run.
   */
  truncation_strategy?: RunCreateParams.TruncationStrategy | null;
}

export namespace RunCreateParams {
  export interface AdditionalMessage {
    /**
     * The text contents of the message.
     */
    content: string | Array<MessagesAPI.MessageContentPartParam>;

    /**
     * The role of the entity that is creating the message. Allowed values include:
     *
     * - `user`: Indicates the message is sent by an actual user and should be used in
     *   most cases to represent user-generated messages.
     * - `assistant`: Indicates the message is generated by the assistant. Use this
     *   value to insert messages from the assistant into the conversation.
     */
    role: 'user' | 'assistant';

    /**
     * A list of files attached to the message, and the tools they should be added to.
     */
    attachments?: Array<AdditionalMessage.Attachment> | null;

    /**
     * Set of 16 key-value pairs that can be attached to an object. This can be useful
     * for storing additional information about the object in a structured format. Keys
     * can be a maximum of 64 characters long and values can be a maxium of 512
     * characters long.
     */
    metadata?: unknown | null;
  }

  export namespace AdditionalMessage {
    export interface Attachment {
      /**
       * The ID of the file to attach to the message.
       */
      file_id?: string;

      /**
       * The tools to add this file to.
       */
      tools?: Array<AssistantsAPI.CodeInterpreterTool | Attachment.FileSearch>;
    }

    export namespace Attachment {
      export interface FileSearch {
        /**
         * The type of tool being defined: `file_search`
         */
        type: 'file_search';
      }
    }
  }

  /**
   * Controls for how a thread will be truncated prior to the run. Use this to
   * control the intial context window of the run.
   */
  export interface TruncationStrategy {
    /**
     * The truncation strategy to use for the thread. The default is `auto`. If set to
     * `last_messages`, the thread will be truncated to the n most recent messages in
     * the thread. When set to `auto`, messages in the middle of the thread will be
     * dropped to fit the context length of the model, `max_prompt_tokens`.
     */
    type: 'auto' | 'last_messages';

    /**
     * The number of most recent messages from the thread when constructing the context
     * for the run.
     */
    last_messages?: number | null;
  }

  export type RunCreateParamsNonStreaming = RunsAPI.RunCreateParamsNonStreaming;
  export type RunCreateParamsStreaming = RunsAPI.RunCreateParamsStreaming;
}

export interface RunCreateParamsNonStreaming extends RunCreateParamsBase {
  /**
   * Body param: If `true`, returns a stream of events that happen during the Run as
   * server-sent events, terminating when the Run enters a terminal state with a
   * `data: [DONE]` message.
   */
  stream?: false | null;
}

export interface RunCreateParamsStreaming extends RunCreateParamsBase {
  /**
   * Body param: If `true`, returns a stream of events that happen during the Run as
   * server-sent events, terminating when the Run enters a terminal state with a
   * `data: [DONE]` message.
   */
  stream: true;
}

export interface RunUpdateParams {
  /**
   * Set of 16 key-value pairs that can be attached to an object. This can be useful
   * for storing additional information about the object in a structured format. Keys
   * can be a maximum of 64 characters long and values can be a maxium of 512
   * characters long.
   */
  metadata?: unknown | null;
}

export interface RunListParams extends CursorPageParams {
  /**
   * A cursor for use in pagination. `before` is an object ID that defines your place
   * in the list. For instance, if you make a list request and receive 100 objects,
   * starting with obj_foo, your subsequent call can include before=obj_foo in order
   * to fetch the previous page of the list.
   */
  before?: string;

  /**
   * Sort order by the `created_at` timestamp of the objects. `asc` for ascending
   * order and `desc` for descending order.
   */
  order?: 'asc' | 'desc';
}

export interface RunCreateAndPollParams {
  /**
   * The ID of the
   * [assistant](https://platform.openai.com/docs/api-reference/assistants) to use to
   * execute this run.
   */
  assistant_id: string;

  /**
   * Appends additional instructions at the end of the instructions for the run. This
   * is useful for modifying the behavior on a per-run basis without overriding other
   * instructions.
   */
  additional_instructions?: string | null;

  /**
   * Adds additional messages to the thread before creating the run.
   */
  additional_messages?: Array<RunCreateAndPollParams.AdditionalMessage> | null;

  /**
   * Overrides the
   * [instructions](https://platform.openai.com/docs/api-reference/assistants/createAssistant)
   * of the assistant. This is useful for modifying the behavior on a per-run basis.
   */
  instructions?: string | null;

  /**
   * The maximum number of completion tokens that may be used over the course of the
   * run. The run will make a best effort to use only the number of completion tokens
   * specified, across multiple turns of the run. If the run exceeds the number of
   * completion tokens specified, the run will end with status `incomplete`. See
   * `incomplete_details` for more info.
   */
  max_completion_tokens?: number | null;

  /**
   * The maximum number of prompt tokens that may be used over the course of the run.
   * The run will make a best effort to use only the number of prompt tokens
   * specified, across multiple turns of the run. If the run exceeds the number of
   * prompt tokens specified, the run will end with status `incomplete`. See
   * `incomplete_details` for more info.
   */
  max_prompt_tokens?: number | null;

  /**
   * Set of 16 key-value pairs that can be attached to an object. This can be useful
   * for storing additional information about the object in a structured format. Keys
   * can be a maximum of 64 characters long and values can be a maxium of 512
   * characters long.
   */
  metadata?: unknown | null;

  /**
   * The ID of the [Model](https://platform.openai.com/docs/api-reference/models) to
   * be used to execute this run. If a value is provided here, it will override the
   * model associated with the assistant. If not, the model associated with the
   * assistant will be used.
   */
  model?:
    | (string & {})
    | 'gpt-4o'
    | 'gpt-4o-2024-05-13'
    | 'gpt-4-turbo'
    | 'gpt-4-turbo-2024-04-09'
    | 'gpt-4-0125-preview'
    | 'gpt-4-turbo-preview'
    | 'gpt-4-1106-preview'
    | 'gpt-4-vision-preview'
    | 'gpt-4'
    | 'gpt-4-0314'
    | 'gpt-4-0613'
    | 'gpt-4-32k'
    | 'gpt-4-32k-0314'
    | 'gpt-4-32k-0613'
    | 'gpt-3.5-turbo'
    | 'gpt-3.5-turbo-16k'
    | 'gpt-3.5-turbo-0613'
    | 'gpt-3.5-turbo-1106'
    | 'gpt-3.5-turbo-0125'
    | 'gpt-3.5-turbo-16k-0613'
    | null;

  /**
   * Specifies the format that the model must output. Compatible with
   * [GPT-4o](https://platform.openai.com/docs/models/gpt-4o),
   * [GPT-4 Turbo](https://platform.openai.com/docs/models/gpt-4-turbo-and-gpt-4),
   * and all GPT-3.5 Turbo models since `gpt-3.5-turbo-1106`.
   *
   * Setting to `{ "type": "json_object" }` enables JSON mode, which guarantees the
   * message the model generates is valid JSON.
   *
   * **Important:** when using JSON mode, you **must** also instruct the model to
   * produce JSON yourself via a system or user message. Without this, the model may
   * generate an unending stream of whitespace until the generation reaches the token
   * limit, resulting in a long-running and seemingly "stuck" request. Also note that
   * the message content may be partially cut off if `finish_reason="length"`, which
   * indicates the generation exceeded `max_tokens` or the conversation exceeded the
   * max context length.
   */
  response_format?: ThreadsAPI.AssistantResponseFormatOption | null;

  /**
   * What sampling temperature to use, between 0 and 2. Higher values like 0.8 will
   * make the output more random, while lower values like 0.2 will make it more
   * focused and deterministic.
   */
  temperature?: number | null;

  /**
   * Controls which (if any) tool is called by the model. `none` means the model will
   * not call any tools and instead generates a message. `auto` is the default value
   * and means the model can pick between generating a message or calling one or more
   * tools. `required` means the model must call one or more tools before responding
   * to the user. Specifying a particular tool like `{"type": "file_search"}` or
   * `{"type": "function", "function": {"name": "my_function"}}` forces the model to
   * call that tool.
   */
  tool_choice?: ThreadsAPI.AssistantToolChoiceOption | null;

  /**
   * Override the tools the assistant can use for this run. This is useful for
   * modifying the behavior on a per-run basis.
   */
  tools?: Array<AssistantsAPI.AssistantTool> | null;

  /**
   * An alternative to sampling with temperature, called nucleus sampling, where the
   * model considers the results of the tokens with top_p probability mass. So 0.1
   * means only the tokens comprising the top 10% probability mass are considered.
   *
   * We generally recommend altering this or temperature but not both.
   */
  top_p?: number | null;

  /**
   * Controls for how a thread will be truncated prior to the run. Use this to
   * control the intial context window of the run.
   */
  truncation_strategy?: RunCreateAndPollParams.TruncationStrategy | null;
}

export namespace RunCreateAndPollParams {
  export interface AdditionalMessage {
    /**
     * The text contents of the message.
     */
    content: string | Array<MessagesAPI.MessageContentPartParam>;

    /**
     * The role of the entity that is creating the message. Allowed values include:
     *
     * - `user`: Indicates the message is sent by an actual user and should be used in
     *   most cases to represent user-generated messages.
     * - `assistant`: Indicates the message is generated by the assistant. Use this
     *   value to insert messages from the assistant into the conversation.
     */
    role: 'user' | 'assistant';

    /**
     * A list of files attached to the message, and the tools they should be added to.
     */
    attachments?: Array<AdditionalMessage.Attachment> | null;

    /**
     * Set of 16 key-value pairs that can be attached to an object. This can be useful
     * for storing additional information about the object in a structured format. Keys
     * can be a maximum of 64 characters long and values can be a maxium of 512
     * characters long.
     */
    metadata?: unknown | null;
  }

  export namespace AdditionalMessage {
    export interface Attachment {
      /**
       * The ID of the file to attach to the message.
       */
      file_id?: string;

      /**
       * The tools to add this file to.
       */
      tools?: Array<AssistantsAPI.CodeInterpreterTool | AssistantsAPI.FileSearchTool>;
    }
  }

  /**
   * Controls for how a thread will be truncated prior to the run. Use this to
   * control the intial context window of the run.
   */
  export interface TruncationStrategy {
    /**
     * The truncation strategy to use for the thread. The default is `auto`. If set to
     * `last_messages`, the thread will be truncated to the n most recent messages in
     * the thread. When set to `auto`, messages in the middle of the thread will be
     * dropped to fit the context length of the model, `max_prompt_tokens`.
     */
    type: 'auto' | 'last_messages';

    /**
     * The number of most recent messages from the thread when constructing the context
     * for the run.
     */
    last_messages?: number | null;
  }
}

export interface RunCreateAndStreamParams {
  /**
   * The ID of the
   * [assistant](https://platform.openai.com/docs/api-reference/assistants) to use to
   * execute this run.
   */
  assistant_id: string;

  /**
   * Appends additional instructions at the end of the instructions for the run. This
   * is useful for modifying the behavior on a per-run basis without overriding other
   * instructions.
   */
  additional_instructions?: string | null;

  /**
   * Adds additional messages to the thread before creating the run.
   */
  additional_messages?: Array<RunCreateAndStreamParams.AdditionalMessage> | null;

  /**
   * Overrides the
   * [instructions](https://platform.openai.com/docs/api-reference/assistants/createAssistant)
   * of the assistant. This is useful for modifying the behavior on a per-run basis.
   */
  instructions?: string | null;

  /**
   * The maximum number of completion tokens that may be used over the course of the
   * run. The run will make a best effort to use only the number of completion tokens
   * specified, across multiple turns of the run. If the run exceeds the number of
   * completion tokens specified, the run will end with status `incomplete`. See
   * `incomplete_details` for more info.
   */
  max_completion_tokens?: number | null;

  /**
   * The maximum number of prompt tokens that may be used over the course of the run.
   * The run will make a best effort to use only the number of prompt tokens
   * specified, across multiple turns of the run. If the run exceeds the number of
   * prompt tokens specified, the run will end with status `incomplete`. See
   * `incomplete_details` for more info.
   */
  max_prompt_tokens?: number | null;

  /**
   * Set of 16 key-value pairs that can be attached to an object. This can be useful
   * for storing additional information about the object in a structured format. Keys
   * can be a maximum of 64 characters long and values can be a maxium of 512
   * characters long.
   */
  metadata?: unknown | null;

  /**
   * The ID of the [Model](https://platform.openai.com/docs/api-reference/models) to
   * be used to execute this run. If a value is provided here, it will override the
   * model associated with the assistant. If not, the model associated with the
   * assistant will be used.
   */
  model?:
    | (string & {})
    | 'gpt-4o'
    | 'gpt-4o-2024-05-13'
    | 'gpt-4-turbo'
    | 'gpt-4-turbo-2024-04-09'
    | 'gpt-4-0125-preview'
    | 'gpt-4-turbo-preview'
    | 'gpt-4-1106-preview'
    | 'gpt-4-vision-preview'
    | 'gpt-4'
    | 'gpt-4-0314'
    | 'gpt-4-0613'
    | 'gpt-4-32k'
    | 'gpt-4-32k-0314'
    | 'gpt-4-32k-0613'
    | 'gpt-3.5-turbo'
    | 'gpt-3.5-turbo-16k'
    | 'gpt-3.5-turbo-0613'
    | 'gpt-3.5-turbo-1106'
    | 'gpt-3.5-turbo-0125'
    | 'gpt-3.5-turbo-16k-0613'
    | null;

  /**
   * Specifies the format that the model must output. Compatible with
   * [GPT-4o](https://platform.openai.com/docs/models/gpt-4o),
   * [GPT-4 Turbo](https://platform.openai.com/docs/models/gpt-4-turbo-and-gpt-4),
   * and all GPT-3.5 Turbo models since `gpt-3.5-turbo-1106`.
   *
   * Setting to `{ "type": "json_object" }` enables JSON mode, which guarantees the
   * message the model generates is valid JSON.
   *
   * **Important:** when using JSON mode, you **must** also instruct the model to
   * produce JSON yourself via a system or user message. Without this, the model may
   * generate an unending stream of whitespace until the generation reaches the token
   * limit, resulting in a long-running and seemingly "stuck" request. Also note that
   * the message content may be partially cut off if `finish_reason="length"`, which
   * indicates the generation exceeded `max_tokens` or the conversation exceeded the
   * max context length.
   */
  response_format?: ThreadsAPI.AssistantResponseFormatOption | null;

  /**
   * What sampling temperature to use, between 0 and 2. Higher values like 0.8 will
   * make the output more random, while lower values like 0.2 will make it more
   * focused and deterministic.
   */
  temperature?: number | null;

  /**
   * Controls which (if any) tool is called by the model. `none` means the model will
   * not call any tools and instead generates a message. `auto` is the default value
   * and means the model can pick between generating a message or calling one or more
   * tools. `required` means the model must call one or more tools before responding
   * to the user. Specifying a particular tool like `{"type": "file_search"}` or
   * `{"type": "function", "function": {"name": "my_function"}}` forces the model to
   * call that tool.
   */
  tool_choice?: ThreadsAPI.AssistantToolChoiceOption | null;

  /**
   * Override the tools the assistant can use for this run. This is useful for
   * modifying the behavior on a per-run basis.
   */
  tools?: Array<AssistantsAPI.AssistantTool> | null;

  /**
   * An alternative to sampling with temperature, called nucleus sampling, where the
   * model considers the results of the tokens with top_p probability mass. So 0.1
   * means only the tokens comprising the top 10% probability mass are considered.
   *
   * We generally recommend altering this or temperature but not both.
   */
  top_p?: number | null;

  /**
   * Controls for how a thread will be truncated prior to the run. Use this to
   * control the intial context window of the run.
   */
  truncation_strategy?: RunCreateAndStreamParams.TruncationStrategy | null;
}

export namespace RunCreateAndStreamParams {
  export interface AdditionalMessage {
    /**
     * The text contents of the message.
     */
    content: string | Array<MessagesAPI.MessageContentPartParam>;

    /**
     * The role of the entity that is creating the message. Allowed values include:
     *
     * - `user`: Indicates the message is sent by an actual user and should be used in
     *   most cases to represent user-generated messages.
     * - `assistant`: Indicates the message is generated by the assistant. Use this
     *   value to insert messages from the assistant into the conversation.
     */
    role: 'user' | 'assistant';

    /**
     * A list of files attached to the message, and the tools they should be added to.
     */
    attachments?: Array<AdditionalMessage.Attachment> | null;

    /**
     * Set of 16 key-value pairs that can be attached to an object. This can be useful
     * for storing additional information about the object in a structured format. Keys
     * can be a maximum of 64 characters long and values can be a maxium of 512
     * characters long.
     */
    metadata?: unknown | null;
  }

  export namespace AdditionalMessage {
    export interface Attachment {
      /**
       * The ID of the file to attach to the message.
       */
      file_id?: string;

      /**
       * The tools to add this file to.
       */
      tools?: Array<AssistantsAPI.CodeInterpreterTool | AssistantsAPI.FileSearchTool>;
    }
  }

  /**
   * Controls for how a thread will be truncated prior to the run. Use this to
   * control the intial context window of the run.
   */
  export interface TruncationStrategy {
    /**
     * The truncation strategy to use for the thread. The default is `auto`. If set to
     * `last_messages`, the thread will be truncated to the n most recent messages in
     * the thread. When set to `auto`, messages in the middle of the thread will be
     * dropped to fit the context length of the model, `max_prompt_tokens`.
     */
    type: 'auto' | 'last_messages';

    /**
     * The number of most recent messages from the thread when constructing the context
     * for the run.
     */
    last_messages?: number | null;
  }
}

export interface RunStreamParams {
  /**
   * The ID of the
   * [assistant](https://platform.openai.com/docs/api-reference/assistants) to use to
   * execute this run.
   */
  assistant_id: string;

  /**
   * Appends additional instructions at the end of the instructions for the run. This
   * is useful for modifying the behavior on a per-run basis without overriding other
   * instructions.
   */
  additional_instructions?: string | null;

  /**
   * Adds additional messages to the thread before creating the run.
   */
  additional_messages?: Array<RunStreamParams.AdditionalMessage> | null;

  /**
   * Overrides the
   * [instructions](https://platform.openai.com/docs/api-reference/assistants/createAssistant)
   * of the assistant. This is useful for modifying the behavior on a per-run basis.
   */
  instructions?: string | null;

  /**
   * The maximum number of completion tokens that may be used over the course of the
   * run. The run will make a best effort to use only the number of completion tokens
   * specified, across multiple turns of the run. If the run exceeds the number of
   * completion tokens specified, the run will end with status `incomplete`. See
   * `incomplete_details` for more info.
   */
  max_completion_tokens?: number | null;

  /**
   * The maximum number of prompt tokens that may be used over the course of the run.
   * The run will make a best effort to use only the number of prompt tokens
   * specified, across multiple turns of the run. If the run exceeds the number of
   * prompt tokens specified, the run will end with status `incomplete`. See
   * `incomplete_details` for more info.
   */
  max_prompt_tokens?: number | null;

  /**
   * Set of 16 key-value pairs that can be attached to an object. This can be useful
   * for storing additional information about the object in a structured format. Keys
   * can be a maximum of 64 characters long and values can be a maxium of 512
   * characters long.
   */
  metadata?: unknown | null;

  /**
   * The ID of the [Model](https://platform.openai.com/docs/api-reference/models) to
   * be used to execute this run. If a value is provided here, it will override the
   * model associated with the assistant. If not, the model associated with the
   * assistant will be used.
   */
  model?:
    | (string & {})
    | 'gpt-4o'
    | 'gpt-4o-2024-05-13'
    | 'gpt-4-turbo'
    | 'gpt-4-turbo-2024-04-09'
    | 'gpt-4-0125-preview'
    | 'gpt-4-turbo-preview'
    | 'gpt-4-1106-preview'
    | 'gpt-4-vision-preview'
    | 'gpt-4'
    | 'gpt-4-0314'
    | 'gpt-4-0613'
    | 'gpt-4-32k'
    | 'gpt-4-32k-0314'
    | 'gpt-4-32k-0613'
    | 'gpt-3.5-turbo'
    | 'gpt-3.5-turbo-16k'
    | 'gpt-3.5-turbo-0613'
    | 'gpt-3.5-turbo-1106'
    | 'gpt-3.5-turbo-0125'
    | 'gpt-3.5-turbo-16k-0613'
    | null;

  /**
   * Specifies the format that the model must output. Compatible with
   * [GPT-4o](https://platform.openai.com/docs/models/gpt-4o),
   * [GPT-4 Turbo](https://platform.openai.com/docs/models/gpt-4-turbo-and-gpt-4),
   * and all GPT-3.5 Turbo models since `gpt-3.5-turbo-1106`.
   *
   * Setting to `{ "type": "json_object" }` enables JSON mode, which guarantees the
   * message the model generates is valid JSON.
   *
   * **Important:** when using JSON mode, you **must** also instruct the model to
   * produce JSON yourself via a system or user message. Without this, the model may
   * generate an unending stream of whitespace until the generation reaches the token
   * limit, resulting in a long-running and seemingly "stuck" request. Also note that
   * the message content may be partially cut off if `finish_reason="length"`, which
   * indicates the generation exceeded `max_tokens` or the conversation exceeded the
   * max context length.
   */
  response_format?: ThreadsAPI.AssistantResponseFormatOption | null;

  /**
   * What sampling temperature to use, between 0 and 2. Higher values like 0.8 will
   * make the output more random, while lower values like 0.2 will make it more
   * focused and deterministic.
   */
  temperature?: number | null;

  /**
   * Controls which (if any) tool is called by the model. `none` means the model will
   * not call any tools and instead generates a message. `auto` is the default value
   * and means the model can pick between generating a message or calling one or more
   * tools. `required` means the model must call one or more tools before responding
   * to the user. Specifying a particular tool like `{"type": "file_search"}` or
   * `{"type": "function", "function": {"name": "my_function"}}` forces the model to
   * call th