openapi: 3.1.0
info:
  title: LangSmith access_policies threads API
  description: 'The LangSmith API is used to programmatically create and manage LangSmith resources.


    ## Host

    https://api.smith.langchain.com


    ## Authentication

    To authenticate with the LangSmith API, set the `X-Api-Key` header

    to a valid [LangSmith API key](https://docs.langchain.com/langsmith/create-account-api-key#create-an-api-key).


    '
  version: 0.1.0
servers:
- url: /
tags:
- name: threads
paths:
  /v2/threads/query:
    post:
      security:
      - API Key: []
      - Tenant ID: []
      - Bearer Auth: []
      description: '**Alpha:** The request and response contract may change;

        Query threads within a project (session), with cursor-based pagination.

        Returns threads matching the given time range and optional filter.'
      tags:
      - threads
      summary: Query Threads
      parameters: []
      responses:
        '200':
          description: items and pagination
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/threads.QueryThreadsResponseBody'
        '400':
          description: Bad Request
          content:
            application/json:
              schema:
                type: object
                additionalProperties:
                  type: string
        '403':
          description: Forbidden
          content:
            application/json:
              schema:
                type: object
                additionalProperties:
                  type: string
        '422':
          description: Unprocessable Entity
          content:
            application/json:
              schema:
                type: object
                additionalProperties:
                  type: string
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/threads.QueryThreadsRequestBody'
  /v2/threads/{thread_id}/traces:
    get:
      security:
      - API Key: []
      - Tenant ID: []
      - Bearer Auth: []
      description: '**Alpha:** The request and response contract may change;

        Retrieve all traces belonging to a specific thread within a project.'
      tags:
      - threads
      summary: Query Thread Traces
      parameters:
      - description: Thread ID
        name: thread_id
        in: path
        required: true
        schema:
          type: string
      - description: '`cursor` is the opaque string from a previous response''s `next_cursor`. Omit on the first request; pass the returned cursor to fetch the next page.'
        name: cursor
        in: query
        schema:
          type: string
          title: Cursor
      - description: '`filter` narrows which traces are returned for this thread, using a LangSmith filter expression evaluated against each root trace run.

          For example: eq(status, "success") or has(tags, "production").

          See https://docs.langchain.com/langsmith/trace-query-syntax#filter-query-language for syntax.'
        name: filter
        in: query
        schema:
          type: string
          title: Filter
      - example: 20
        description: '`page_size` is the maximum number of traces to return in this response. Defaults to 20 when omitted; must be between 1 and 100 inclusive when set.'
        name: page_size
        in: query
        schema:
          type: integer
          default: 20
          minimum: 1
          maximum: 100
          title: Page Size
      - description: '`project_id` is the tracing project UUID (required).'
        name: project_id
        in: query
        required: true
        schema:
          type: string
          format: uuid
          title: Project Id
      - example:
        - NAME
        - START_TIME
        description: '`selects` lists which properties to include on each returned trace (repeatable query parameter). Accepts any value of the `ThreadTraceSelectField` enum. Properties not listed are omitted from each trace object; `trace_id` is always returned.'
        name: selects
        in: query
        style: form
        explode: true
        schema:
          type: array
          items:
            enum:
            - THREAD_ID
            - TRACE_ID
            - OP
            - PROMPT_TOKENS
            - COMPLETION_TOKENS
            - TOTAL_TOKENS
            - START_TIME
            - END_TIME
            - LATENCY
            - FIRST_TOKEN_TIME
            - INPUTS_PREVIEW
            - OUTPUTS_PREVIEW
            - PROMPT_COST
            - COMPLETION_COST
            - TOTAL_COST
            - PROMPT_TOKEN_DETAILS
            - COMPLETION_TOKEN_DETAILS
            - PROMPT_COST_DETAILS
            - COMPLETION_COST_DETAILS
            - NAME
            - ERROR_PREVIEW
            type: string
          title: Selects
      responses:
        '200':
          description: items and pagination
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/threads.QueryThreadTracesResponseBody'
        '400':
          description: Bad Request
          content:
            application/json:
              schema:
                type: object
                additionalProperties:
                  type: string
        '403':
          description: Forbidden
          content:
            application/json:
              schema:
                type: object
                additionalProperties:
                  type: string
        '422':
          description: Unprocessable Entity
          content:
            application/json:
              schema:
                type: object
                additionalProperties:
                  type: string
components:
  schemas:
    query.RunFeedbackStats:
      type: object
      additionalProperties:
        $ref: '#/components/schemas/query.RunFeedbackStat'
    threads.ThreadTraceListItem:
      type: object
      properties:
        completion_cost:
          description: '`completion_cost` is the estimated USD cost for the completion. Omitted unless included in `selects`.'
          type: number
        completion_cost_details:
          description: '`completion_cost_details` is the USD cost breakdown for completion-side categories; per-category values are under `raw`. Omitted unless included in `selects`.'
          allOf:
          - $ref: '#/components/schemas/query.RunCompletionCostDetails'
        completion_token_details:
          description: '`completion_token_details` is the completion-side token breakdown by category; per-category counts are under `raw`. Omitted unless included in `selects`.'
          allOf:
          - $ref: '#/components/schemas/query.RunCompletionTokenDetails'
        completion_tokens:
          description: '`completion_tokens` is the completion-side token count. Omitted unless included in `selects`.'
          type: integer
        end_time:
          description: '`end_time` is when the root run ended (RFC3339 date-time). JSON null if the run is still in progress. Omitted unless included in `selects`.'
          type: string
          format: date-time
          example: '2025-01-15T12:00:01.500Z'
        error_preview:
          description: '`error_preview` is a short error summary when the run failed. Omitted unless included in `selects`.'
          type: string
        first_token_time:
          description: '`first_token_time` is when the first output token was produced (RFC3339 date-time), for streamed runs when that metadata exists. Omitted unless included in `selects`.'
          type: string
          format: date-time
          example: '2024-01-15T10:30:00.312Z'
        inputs_preview:
          description: '`inputs_preview` is a truncated text preview of inputs. Omitted unless included in `selects`.'
          type: string
        latency:
          description: '`latency` is wall-clock duration from start to end in seconds. Omitted unless included in `selects`.'
          type: number
        name:
          description: '`name` is a human-readable label for the root run (for example the model name, function name, or step name chosen when the run was traced). Omitted unless included in `selects`.'
          type: string
        op:
          description: '`op` is a numeric code identifying the root run''s `run_type` (for example LLM vs. tool vs. chain). Encoded as a number for compatibility with legacy clients; prefer the string `run_type` on `RunResponse` when available. Omitted unless included in `selects`.'
          type: number
        outputs_preview:
          description: '`outputs_preview` is a truncated text preview of outputs. Omitted unless included in `selects`.'
          type: string
        prompt_cost:
          description: '`prompt_cost` is the estimated USD cost for the prompt. Omitted unless included in `selects`.'
          type: number
        prompt_cost_details:
          description: '`prompt_cost_details` is the USD cost breakdown for prompt-side categories; per-category values are under `raw`. Omitted unless included in `selects`.'
          allOf:
          - $ref: '#/components/schemas/query.RunPromptCostDetails'
        prompt_token_details:
          description: '`prompt_token_details` is the prompt-side token breakdown by category; per-category counts are under nested `raw`. Omitted unless included in `selects`.'
          allOf:
          - $ref: '#/components/schemas/query.RunPromptTokenDetails'
        prompt_tokens:
          description: '`prompt_tokens` is the prompt-side token count. Omitted unless included in `selects`.'
          type: integer
        start_time:
          description: '`start_time` is when the trace started (RFC3339 date-time). Omitted unless included in `selects`.'
          type: string
          format: date-time
          example: '2025-01-15T12:00:00.000Z'
        thread_id:
          description: '`thread_id` is the conversation thread UUID that contains this trace. Matches the `thread_id` path parameter of the request. Omitted unless included in `selects`.'
          type: string
          format: uuid
          example: d4e5f6a7-b8c9-4d5e-1f2a-3b4c5d6e7f8a
        total_cost:
          description: '`total_cost` is the estimated total USD cost for the root run. Omitted unless included in `selects`.'
          type: number
        total_tokens:
          description: '`total_tokens` is the total token count (prompt plus completion). Omitted unless included in `selects`.'
          type: integer
        trace_id:
          description: '`trace_id` is the UUID of this trace (the root run). Always present.'
          type: string
          format: uuid
          example: 018e4c7e-a9fb-7ef0-a5b6-6ea3a82e9327
    query.RunPromptTokenDetails:
      type: object
      properties:
        raw:
          description: '`raw` maps each category name to its prompt-token count.'
          type: object
          additionalProperties:
            type: integer
            format: int64
    threads.QueryThreadsResponseBody:
      type: object
      properties:
        has_more:
          description: '`has_more` is true when another page of threads exists after this one.'
          type: boolean
          example: true
        items:
          description: '`items` is the page of thread summaries, sorted by the thread''s most recent activity.'
          type: array
          items:
            $ref: '#/components/schemas/threads.ThreadListItem'
        next_cursor:
          description: '`next_cursor` is the opaque cursor to pass as `cursor` on the next request when `has_more` is true. Omitted on the final page.'
          type: string
          example: eyJydW5zX2N1cnNvciI6Imx0KGN1cnNvciwiLi4uIikifQ==
    query.RunCompletionTokenDetails:
      type: object
      properties:
        raw:
          description: '`raw` maps each category name to its completion-token count.'
          type: object
          additionalProperties:
            type: integer
            format: int64
    threads.ThreadListItem:
      type: object
      properties:
        count:
          description: '`count` is how many root traces (conversation turns) fall in this thread for the query time range.'
          type: integer
          example: 3
        feedback_stats:
          description: '`feedback_stats` is the aggregated feedback across traces in the thread, keyed by feedback key; shape matches `feedback_stats` on a single run.'
          allOf:
          - $ref: '#/components/schemas/query.RunFeedbackStats'
        first_inputs:
          description: '`first_inputs` is a truncated preview of inputs from the earliest trace in the thread for the query window.'
          type: string
        first_trace_id:
          description: '`first_trace_id` is the root trace UUID for the chronologically first trace in the query time window.'
          type: string
          format: uuid
          example: 018e4c7e-a9fb-7ef0-a5b6-6ea3a82e9327
        last_error:
          description: '`last_error` is a short error summary from the most recent failing trace in the thread. Absent when there is no error in the window.'
          type: string
        last_outputs:
          description: '`last_outputs` is a truncated preview of outputs from the latest trace in the thread for the query window.'
          type: string
        last_trace_id:
          description: '`last_trace_id` is the root trace UUID for the chronologically last trace in the query time window.'
          type: string
          format: uuid
          example: 0190a1b2-c3d4-7ef0-a5b6-6ea3a82e9328
        latency_p50:
          description: '`latency_p50` is the approximate median end-to-end latency of traces in the thread, in seconds.'
          type: number
          example: 0.15
        latency_p99:
          description: '`latency_p99` is the approximate 99th percentile end-to-end latency of traces in the thread, in seconds.'
          type: number
          example: 0.42
        max_start_time:
          description: '`max_start_time` is the latest trace start time in the thread (RFC3339 date-time).'
          type: string
          format: date-time
          example: '2025-01-15T12:05:00.000Z'
        min_start_time:
          description: '`min_start_time` is the earliest trace start time in the thread (RFC3339 date-time).'
          type: string
          format: date-time
          example: '2025-01-15T12:00:00.000Z'
        start_time:
          description: '`start_time` is a reference start time for this row (RFC3339 date-time), such as for sorting.'
          type: string
          format: date-time
          example: '2025-01-15T12:00:00.000Z'
        thread_id:
          description: '`thread_id` identifies this conversation thread within the project from the request body `project_id`.'
          type: string
          format: uuid
          example: 018e4c7e-a9fb-7ef0-a5b6-6ea3a82e9327
        total_cost:
          description: '`total_cost` is the sum of estimated USD cost across those traces.'
          type: number
          example: 0.045
        total_cost_details:
          description: '`total_cost_details` sums per-category estimated USD cost across traces in the thread. Keys mirror `total_token_details`.


            Example: `{"cache_read": 0.012, "reasoning": 0.008}`.'
          type: object
          additionalProperties:
            type: number
            format: float64
        total_token_details:
          description: '`total_token_details` sums per-category token counts across traces in the thread. Keys are model-specific category names (for example `cache_read`, `cache_write`, `reasoning`, `audio`).


            Example: `{"cache_read": 400, "reasoning": 120}`.'
          type: object
          additionalProperties:
            type: integer
            format: int64
        total_tokens:
          description: '`total_tokens` is the sum of token usage across those traces.'
          type: integer
          example: 450
        trace_id:
          description: '`trace_id` is a representative root trace UUID when the summary includes one, for example for deep links.'
          type: string
          format: uuid
          example: 018e4c7e-a9fb-7ef0-a5b6-6ea3a82e9328
    query.RunFeedbackStat:
      type: object
      properties:
        avg:
          description: '`avg` is the arithmetic mean of numeric feedback scores for this key on the run, or `null` when no numeric score has been recorded (for example purely categorical feedback).'
          type: number
          example: 0.87
        comments:
          description: '`comments` is a sample of human-readable comments attached to feedback points for this key, in no particular order. May be empty; is not exhaustive when many comments exist.'
          type: array
          items:
            type: string
          example:
          - good answer
          - needs citation
        contains_thread_feedback:
          description: '`contains_thread_feedback` is true when at least one feedback point for this key was submitted at the thread level (rather than at an individual run). Always false on responses that already describe a single run in isolation.'
          type: boolean
          example: false
        errors:
          description: '`errors` is the number of feedback points recorded as errors rather than successful scores (for example an automated evaluator that raised an exception). Defaults to 0 when no errors occurred.'
          type: integer
          default: 0
          example: 0
        max:
          description: '`max` is the largest numeric feedback score recorded for this key on the run, or `null` when no numeric score has been recorded.'
          type: number
          example: 0.95
        min:
          description: '`min` is the smallest numeric feedback score recorded for this key on the run, or `null` when no numeric score has been recorded.'
          type: number
          example: 0.8
        n:
          description: '`n` is the number of feedback points recorded for this key on the run. For numeric feedback this is the sample size behind `avg`, `min`, `max`, and `stdev`; for categorical feedback it is the sum of the `values` counts.'
          type: integer
          example: 42
        sources:
          description: '`sources` is a sample of feedback sources for this key. Each entry is either a plain string identifier (for example `"api"`, `"app"`, `"model"`) or a JSON object describing a synthetic source (for example `{"type": "__ls_composite_feedback"}` for a computed aggregate). Clients must tolerate both shapes.'
          type: array
          items: {}
        stdev:
          description: '`stdev` is the sample standard deviation of numeric feedback scores for this key on the run, or `null` when it cannot be computed (for example fewer than two numeric scores, or purely categorical feedback).'
          type: number
          example: 0.05
        values:
          description: '`values` is the distribution of categorical feedback labels for this key, mapping each label to its occurrence count. Empty (`{}`) for purely numeric feedback.'
          type: object
          additionalProperties:
            type: integer
            format: int64
    threads.QueryThreadsRequestBody:
      type: object
      properties:
        cursor:
          description: '`cursor` is the opaque string from a previous response''s `next_cursor`. Omit on the first request; pass the returned cursor to fetch the next page.'
          type: string
        filter:
          description: '`filter` narrows which threads are returned, using a LangSmith filter expression evaluated against each thread''s root run.

            For example: has(tags, "production") or eq(status, "error").

            See https://docs.langchain.com/langsmith/trace-query-syntax#filter-query-language for syntax.'
          type: string
        max_start_time:
          description: '`max_start_time` is the inclusive upper bound on thread activity (RFC3339 date-time).'
          type: string
          format: date-time
        min_start_time:
          description: '`min_start_time` is the inclusive lower bound on thread activity (RFC3339 date-time).'
          type: string
          format: date-time
        page_size:
          description: '`page_size` is the maximum number of threads to return in this response. Defaults to 20 when omitted; must be between 1 and 100 inclusive when set. The response may contain fewer threads than `page_size` even when `has_more` is true.'
          type: integer
          default: 20
          maximum: 100
          minimum: 1
          example: 20
        project_id:
          description: '`project_id` is the tracing project UUID.'
          type: string
          format: uuid
          example: 0190a1b2-c3d4-7ef0-a5b6-6ea3a82e9328
    threads.QueryThreadTracesResponseBody:
      type: object
      properties:
        has_more:
          description: '`has_more` is true when another page of traces exists after this one.'
          type: boolean
          example: true
        items:
          description: '`items` is the page of root traces in this thread. Which properties are populated on each trace depends on the `selects` query parameter.'
          type: array
          items:
            $ref: '#/components/schemas/threads.ThreadTraceListItem'
        next_cursor:
          description: '`next_cursor` is the opaque cursor to pass as `cursor` on the next request when `has_more` is true. Omitted on the final page.'
          type: string
          example: eyJydW5zX2N1cnNvciI6Imx0KGN1cnNvciwiLi4uIikifQ==
    query.RunPromptCostDetails:
      type: object
      properties:
        raw:
          description: '`raw` maps each category name to its estimated USD cost.'
          type: object
          additionalProperties:
            type: number
            format: float64
    query.RunCompletionCostDetails:
      type: object
      properties:
        raw:
          description: '`raw` maps each category name to its estimated USD cost.'
          type: object
          additionalProperties:
            type: number
            format: float64
  securitySchemes:
    API Key:
      type: apiKey
      in: header
      name: X-API-Key
    Tenant ID:
      type: apiKey
      in: header
      name: X-Tenant-Id
    Bearer Auth:
      type: http
      description: Bearer tokens are used to authenticate from the UI. Must also specify x-tenant-id or x-organization-id (for org scoped apis).
      scheme: bearer
    Organization ID:
      type: apiKey
      in: header
      name: X-Organization-Id