Skip to main content
POST
/
v1
/
monitors
Python
from parallel import Parallel

client = Parallel()

monitor = client.monitor.create(
    type="event_stream",
    frequency="1d",
    settings={"query": "Extract recent news about AI"},
)
print(monitor.monitor_id)
{
  "type": "event_stream",
  "monitor_id": "<string>",
  "status": "active",
  "frequency": "<string>",
  "processor": "lite",
  "created_at": "2023-11-07T05:31:56Z",
  "settings": {
    "query": "<string>",
    "output_schema": {
      "json_schema": {},
      "type": "json"
    },
    "include_backfill": true,
    "advanced_settings": {
      "source_policy": {
        "exclude_domains": [
          "reddit.com",
          "x.com",
          ".ai"
        ],
        "include_domains": [
          "wikipedia.org",
          "usa.gov",
          ".edu"
        ]
      },
      "location": "us"
    }
  },
  "webhook": {
    "url": "<string>",
    "event_types": [
      "monitor.event.detected"
    ]
  },
  "metadata": {
    "slack_thread_id": "1234567890.123456",
    "user_id": "U123ABC"
  },
  "last_run_at": "2025-01-15T10:30:00Z",
  "output": {
    "latest_snapshot": {
      "basis": [
        {
          "field": "<string>",
          "reasoning": "<string>",
          "citations": [],
          "confidence": "low"
        }
      ],
      "type": "<string>",
      "content": "<string>",
      "mcp_tool_calls": [
        {
          "tool_call_id": "<string>",
          "server_name": "<string>",
          "tool_name": "<string>",
          "arguments": "<string>",
          "content": "<string>",
          "error": "<string>"
        }
      ],
      "beta_fields": {}
    }
  }
}

Documentation Index

Fetch the complete documentation index at: https://docs.parallel.ai/llms.txt

Use this file to discover all available pages before exploring further.

Authorizations

​
x-api-key
string
header
required

Body

application/json

Request body to create a monitor.

The type field at the root determines the expected shape of settings: event_stream requires MonitorEventStreamSettings, and snapshot requires MonitorSnapshotSettings.

​
type
enum<string>
required

Type of monitor to create. event_stream monitors a search query for material changes; snapshot monitors a specific task run's output. Determines the expected shape of settings.

Available options:
event_stream,
snapshot
Examples:

"event_stream"

"snapshot"

​
frequency
string
required

Frequency of the monitor. Format: '' where unit is 'h' (hours), 'd' (days), or 'w' (weeks). Must be between 1h and 30d (inclusive).

Examples:

"1h"

"12h"

"1d"

"7d"

"30d"

​
settings
MonitorEventStreamSettings · object
required

Type-specific settings for an event_stream monitor.

​
processor
enum<string>
default:lite

Processor to use for the monitor. lite is faster and cheaper; base performs more thorough analysis at higher cost and latency. Defaults to lite.

Available options:
lite,
base
​
webhook
MonitorWebhook · object

Webhook to receive notifications about the monitor's execution.

​
metadata
Metadata · object

User-provided metadata stored with the monitor and echoed back in webhook notifications and GET responses, so you can map events to objects in your application. Keys: max 16 chars; values: max 512 chars.

Example:
{
  "slack_thread_id": "1234567890.123456",
  "user_id": "U123ABC"
}

Response

Monitor created successfully.

Response object for a monitor.

The type field at the root determines the concrete shape of settings: event_stream uses MonitorEventStreamResponseSettings, and snapshot uses MonitorSnapshotResponseSettings. Snapshot monitors also carry an output field (MonitorSnapshotOutput) with the latest computed state.

​
type
enum<string>
required

The type of monitor.

Available options:
event_stream,
snapshot
Examples:

"event_stream"

"snapshot"

​
monitor_id
string
required

ID of the monitor.

​
status
enum<string>
required

Status of the monitor.

Available options:
active,
cancelled
Examples:

"active"

"cancelled"

​
frequency
string
required

Frequency of the monitor. Format: '' where unit is 'h' (hours), 'd' (days), or 'w' (weeks). Must be between 1h and 30d (inclusive).

Examples:

"1h"

"12h"

"1d"

"7d"

"30d"

​
processor
enum<string>
required

Processor to use for the monitor. lite is faster and cheaper; base performs more thorough analysis at higher cost and latency. Defaults to lite.

Available options:
lite,
base
Examples:

"lite"

"base"

​
created_at
string<date-time>
required

Timestamp of the creation of the monitor, as an RFC 3339 string.

Example:

"2025-01-15T10:30:00Z"

​
settings
MonitorEventStreamResponseSettings · object
required

Type-specific response fields for an event_stream monitor.

​
webhook
MonitorWebhook · object

Webhook configuration for the monitor.

​
metadata
Metadata · object

User-provided metadata stored with the monitor and echoed back in webhook notifications and GET responses, so you can map events to objects in your application. Keys: max 16 chars; values: max 512 chars.

Example:
{
  "slack_thread_id": "1234567890.123456",
  "user_id": "U123ABC"
}
​
last_run_at
string | null

Timestamp of the last run for the monitor, as an RFC 3339 string.

Example:

"2025-01-15T10:30:00Z"

​
output
MonitorSnapshotOutput · object

Runtime output state. Present only for snapshot monitors; null for event_stream monitors.