Create Research Job

Kick off a Dex Research job.

Persists a research job and starts :class:~dex_api.temporal.research.workflows.research_workflow.ResearchWorkflow on the research task queue.

Scope: one inputs.vector_stores id, or one or more inputs.parse_results ids (no mixing with vector stores). Empty inputs: if the project has exactly one vector store, that store is used automatically (empty-input shortcut).

POST /v1/projects/{project_id}/research

Create Research Job

cURL

curl --request POST \
  --url https://api.example.com/v1/projects/{project_id}/research \
  --header 'Content-Type: application/json' \
  --header 'x-api-key: <api-key>' \
  --header 'x-selected-account-id: <api-key>' \
  --data '
{
  "advanced": {
    "decomposition": {
      "model": "openai/gpt-5.4",
      "prompt": "PROMPT_DECOMPOSITION",
      "reasoning": {
        "effort": "high"
      },
      "temperature": 0
    },
    "subtasks": {
      "model": "openai/gpt-5.3-codex",
      "prompt": "PROMPT_SUBTASKS",
      "temperature": 0
    }
  },
  "inputs": {
    "parse_results": [
      "pres_9f86d081884c7d659a2feaa0c55ad015"
    ],
    "vector_stores": [
      "vector_store_9f86d081884c7d659a2feaa0c55ad015"
    ]
  },
  "model": "openai/gpt-5.4",
  "task": {
    "description": "What is the Commencement Date of the document?",
    "output_schema": {
      "properties": {
        "commencement_date": {
          "description": "The Commencement Date",
          "type": "string"
        }
      },
      "type": "object"
    },
    "plan": "1. Find the Start Date\n2. Search for Commencement Date"
  },
  "tools": {
    "lexical_search": {
      "config": {},
      "enabled": false
    },
    "similarity_search": {
      "config": {
        "filters": {
          "file_id": {
            "$eq": "file_9f86d081884c7d659a2feaa0c55ad015"
          }
        },
        "top_k": 5
      },
      "enabled": true
    }
  }
}
'

202

{
  "id": "<string>",
  "project_id": "<string>",
  "created_at": "2023-11-07T05:31:56Z",
  "object": "job",
  "source_id": "<string>",
  "correlation_id": "<string>",
  "started_at": "2023-11-07T05:31:56Z",
  "completed_at": "2023-11-07T05:31:56Z",
  "result": {},
  "progress": {
    "total": 123,
    "succeeded": 123,
    "failed": 123,
    "cancelled": 123,
    "pending": 123,
    "child_jobs": [
      {
        "source_document_id": "<string>",
        "job_id": "<string>",
        "parse_result_id": "<string>",
        "error": "<string>"
      }
    ]
  },
  "error": "<string>",
  "history": [
    {
      "step": "<string>",
      "timestamp": "2023-11-07T05:31:56Z",
      "duration_ms": 123,
      "status": "<string>",
      "details": {}
    }
  ]
}

Authorizations

x-api-key · string · header · required

API key for authentication

x-selected-account-id · string · header · required

Selected Account ID

Path Parameters

project_id · string · required

Body

application/json

Kickoff body for POST /v1/projects/{project_id}/research.

Use tools for per-tool enablement and configuration (same map shape as :class:~dex_core.models.extraction.ExtractionParameters.agentic_tools on extract jobs).

task · ResearchTask · object · required

User task: goal, optional plan, and structured or report-shaped output.

Show child attributes

model · string · required

Default model for all agent steps unless overridden in advanced

inputs · ResearchInputs · object

Explicit scope. Empty lists: the API resolves to the project's sole vector store when exactly one exists; otherwise set vector_stores or parse_results explicitly.

Show child attributes

tools · Tools · object

Map tool name → enablement + config

Show child attributes

advanced · ResearchAdvanced · object

Per-stage model/prompt/temperature/reasoning overrides

Show child attributes

execute_dependent_subtask_tool_hints · boolean · default: false

When True, DECOMPOSE tool hints for dependent subtasks are executed as warm-start searches (same semantics as extract execute_dependent_subtask_tool_hints). Default False (intent-only hints for dependents).

Response

202

application/json

Successful Response

Job response model representing an asynchronous operation.

id · string · required

ID of the entity

project_id · string · required

ID of the project

operation · enum<string> · required

Operation type (e.g., 'parse')

Available options: parse, batch_parse, extract, research, vector_store, chunk, summarization, create_index, update_index

status · enum<string> · required

Current job status

Available options: pending, running, succeeded, partially_succeeded, failed, cancelled

created_at · string<date-time> · required

When the job was created

object · string · default: job

Allowed value: "job"

source_id · string | null

Source document/file ID

correlation_id · string | null

Request correlation ID for tracing

started_at · string<date-time> | null

When the job started processing

completed_at · string<date-time> | null

When the job completed

result · Result · object

Job result payload when completed

progress · BatchParseProgress · object

Live progress payload (used by batch jobs)

Show child attributes

error · string | null

Error message if job failed

history · JobHistoryEvent · object[] | null

Timeline of job execution events

Show child attributes