Skip to main content
POST
/
v1
/
generation
/
deep_research
curl --request POST \
  --url https://api.weryai.com/v1/generation/deep_research \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '
{
  "mode": "DEEP_RESEARCH",
  "input": "Analyze the latest trends in AI-powered code generation tools",
  "thinking_summaries": "auto",
  "collaborative_planning": true,
  "visualization": "auto",
  "tools": [
    {
      "type": "GOOGLE_SEARCH"
    },
    {
      "type": "FILE_SEARCH",
      "file_search_store_names": [
        "my-corpus-store"
      ]
    }
  ],
  "webhook_url": "https://your-server.com/webhook"
}
'
{
  "status": 0,
  "desc": "success",
  "message": "success",
  "data": {
    "interaction_id": "dr_interaction_abc123",
    "task_id": "task_abc123",
    "status": "processing"
  }
}

Authorizations

​
Authorization
string
header
required

Authenticate using Bearer token. Get your API Key from the WeryAI Console.

Example: Authorization: Bearer sk-xxxxxxxxxxxxxxxx

Body

application/json

Request parameters for submitting a Deep Research task. At least one of input or input_resource must be provided. Exactly one of mode or model must be specified — they are mutually exclusive.

​
mode
enum<string>

Research run mode. Mutually exclusive with model — provide one or the other, not both.

Available options:
DEEP_RESEARCH,
DEEP_RESEARCH_MAX
Example:

"DEEP_RESEARCH"

​
model
string

Model key for the research task. Mutually exclusive with mode — provide one or the other, not both. When model is specified, the task runs synchronously and the result is returned directly in the response data.deep_research_outputs field.

Model NameModel Key
Gemini-2.5-ProGEMINI_25_PRO
Gemini-2.5-FlashGEMINI_25_FLASH
Gemini-3.1-Flash-LiteGEMINI_3_1_FLASH_LITE
Gemini-3.1-ProGEMINI_3_1_PRO
Example:

"GEMINI_25_PRO"

​
previous_interaction_id
string

Interaction ID from a previous research or execution session, used to continue a multi-round conversation.

Example:

"dr_interaction_abc123"

​
thinking_summaries
string

Controls how the model exposes its reasoning chain. Set to auto to let the model decide.

Example:

"auto"

​
collaborative_planning
boolean

When true, the task enters the research/planning phase before execution.

Example:

true

​
input
string

Plain text input for the research task. At least one of input or input_resource must be provided.

Example:

"Analyze the latest trends in AI-powered code generation tools"

​
input_resource
object[]

Structured input list supporting text, image, and document items. At least one of input or input_resource must be provided.

​
visualization
string

Controls whether the agent may generate charts or diagrams. Set to auto to let the model decide.

Example:

"auto"

​
tools
object[]

List of agent tools to enable for this task.

​
webhook_url
string<uri>

Callback URL where results will be sent when the task completes.

Example:

"https://your-server.com/webhook"

​
caller_id
integer<int64>

Caller ID (for business association)

Response

Task submitted successfully

​
status
integer

Business status code (0 indicates success)

Example:

0

​
desc
string

Status description

Example:

"success"

​
message
string

Multi-language message

Example:

"success"

​
data
object