---
canonical_url: "https://docs.composio.dev/toolkits/seltz"
markdown_url: "https://docs.composio.dev/toolkits/seltz.md"
product: "platform"
version: null
last_updated: null
---

# Seltz

Seltz provides web search, cited answers, URL fetching, research agents, and recurring monitors over web, news, people, company, Wikipedia, and financial-report data.

- **Category:** ai web scraping
- **Auth:** API_KEY
- **Composio-managed OAuth available?** N/A
- **Tools:** 16
- **Triggers:** 0
- **Slug:** `SELTZ`
- **Version:** 00000000_00

## Tools

### PAID Answer ($6/1,000 Base, $7/1,000 Pro)

**Slug:** `SELTZ_ANSWER`

PAID: generate one buffered JSON grounded answer with citations for $6/1,000 base or $7/1,000 pro responses; this tool never requests SSE streaming.

#### Input Parameters

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `model` | string ("seltz-base" | "seltz-pro") | No | Answer model and price: seltz-base is $6/1,000 responses; seltz-pro is $7/1,000. |
| `query` | string | Yes | Natural-language question to answer from the selected Seltz index. |
| `scope` | string ("web" | "news" | "wikipedia" | "people" | "companies" | "financial-reports") | No | Data index used to ground the answer. Omitted, empty, or whitespace-only values use web. |
| `system_prompt` | string | No | Optional presentation instructions for tone, voice, or format; grounding and citation rules still apply. |
| `include_content` | boolean | No | Include source document content in each citation when available. |
| `response_format` | string | No | Optional OpenAI-style output format. Choose text, json_object, or json_schema; json_schema requires a name, schema, and strict setting. |

#### Output

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `data` | string | Yes | Data from the action execution |
| `error` | string | No | Error if any occurred during the execution of the action |
| `successful` | boolean | Yes | Whether or not the action execution was successful or not |

### PAID Create Agent Run ($0.25 Low, $0.50 Medium, $1.00 High, $2.00 Max)

**Slug:** `SELTZ_CREATE_AGENT_RUN`

PAID: start an asynchronous multi-step web research run for $0.25 low, $0.50 medium, $1.00 high, or $2.00 max per run. The tool defaults to low for cost safety and returns a run ID to poll with Get Agent Run.

#### Input Parameters

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `query` | string | Yes | Natural-language research question and instructions. |
| `effort` | string ("low" | "medium" | "high" | "max") | No | Research depth and per-run price: low $0.25, medium $0.50, high $1.00, or max $2.00. Defaults to low for cost safety. |
| `output_schema` | string | No | Optional OpenAI-style output format. Choose text, json_object, or json_schema; json_schema requires a name, schema, and strict setting. |

#### Output

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `data` | string | Yes | Data from the action execution |
| `error` | string | No | Error if any occurred during the execution of the action |
| `successful` | boolean | Yes | Whether or not the action execution was successful or not |

### PAID Create Monitor ($3/1,000 Base Checks, $5/1,000 Pro Checks)

**Slug:** `SELTZ_CREATE_MONITOR`

PAID: create a recurring search monitor, safely paused by default; active checks cost $3/1,000 base or $5/1,000 pro checks, and choosing active starts that recurring paid behavior.

#### Input Parameters

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `name` | string | Yes | Unique name among the organization's live monitors, at most 512 UTF-8 bytes. |
| `status` | string ("paused" | "active") | No | Create paused to avoid starting paid checks, or explicitly choose active to begin recurring checks immediately. |
| `cadence` | string | Yes | Single-unit interval from 1 hour through 366 days, such as 3600 (bare seconds), 1h, 12h, 7d, or 1w; compound durations are invalid. |
| `webhook` | object | No | Optional callback configuration for completed or failed monitor runs. |
| `search_requests` | array | Yes | Distinct searches run on every paid monitor check. |

#### Output

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `data` | string | Yes | Data from the action execution |
| `error` | string | No | Error if any occurred during the execution of the action |
| `successful` | boolean | Yes | Whether or not the action execution was successful or not |

### DESTRUCTIVE Cancel Agent Run

**Slug:** `SELTZ_DESTRUCTIVE_CANCEL_AGENT_RUN`

DESTRUCTIVE: best-effort cancellation of a pending or running paid research run. Cancellation can permanently stop work before output exists; a run that is already terminal is returned unchanged, making retries safe.

#### Input Parameters

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `run_id` | string | Yes | Agent run ID to cancel. |

#### Output

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `data` | string | Yes | Data from the action execution |
| `error` | string | No | Error if any occurred during the execution of the action |
| `successful` | boolean | Yes | Whether or not the action execution was successful or not |

### DESTRUCTIVE Delete Monitor

**Slug:** `SELTZ_DESTRUCTIVE_DELETE_MONITOR`

DESTRUCTIVE: permanently remove a monitor from API visibility, immediately hide all of its records, and release its name; the operation is not reversible to callers.

#### Input Parameters

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `monitor_id` | string | Yes | Monitor ID to permanently remove from API visibility. |

#### Output

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `data` | string | Yes | Data from the action execution |
| `error` | string | No | Error if any occurred during the execution of the action |
| `successful` | boolean | Yes | Whether or not the action execution was successful or not |

### DESTRUCTIVE PAID Update Monitor ($3/1,000 Base, $5/1,000 Pro Checks)

**Slug:** `SELTZ_DESTRUCTIVE_UPDATE_MONITOR`

DESTRUCTIVE PAID: update a monitor with last-writer-wins semantics. No update-call price is documented, but activating or changing active searches or cadence controls recurring checks at $3/1,000 base or $5/1,000 pro. Omitted fields remain unchanged.

#### Input Parameters

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `name` | string | No | New unique monitor name; omit to retain the current name. |
| `status` | string ("active" | "paused") | No | Set active to start or resume recurring paid checks, or paused to stop scheduling new checks; omit to retain the current status. |
| `cadence` | string | No | New single-unit cadence from 1 hour through 366 days, such as 3600 (bare seconds), 1h, 12h, 7d, or 1w; omit to retain the current cadence. |
| `webhook` | object | No | Replacement webhook. Resend its URL and events because Seltz replaces the whole webhook; omit to retain the current webhook. |
| `monitor_id` | string | Yes | Monitor ID to update. |
| `remove_webhook` | boolean | No | Set true to explicitly send webhook:null and remove the current webhook. Cannot be combined with webhook. |
| `search_requests` | array | No | Replacement for the entire search list; omit it or pass an empty list to retain the current searches. Each search defaults to the base tier for cost safety. |

#### Output

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `data` | string | Yes | Data from the action execution |
| `error` | string | No | Error if any occurred during the execution of the action |
| `successful` | boolean | Yes | Whether or not the action execution was successful or not |

### PAID Fetch URLs ($10/1,000 Successful URLs)

**Slug:** `SELTZ_FETCH_URLS`

PAID: fetch 1-20 unique web pages concurrently as LLM-ready Markdown for $10/1,000 successfully fetched URLs; failed URLs are not billed and remain per-item errors in the response.

#### Input Parameters

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `urls` | array | Yes | Unique non-empty URL strings to fetch, from 1 to 20 entries, without surrounding whitespace and at most 2,048 UTF-8 bytes each. Use absolute HTTP(S) URLs; malformed, unreachable, or unsupported values are returned as per-item errors rather than rejecting the batch. |
| `timeout_ms` | integer | No | Per-URL timeout in milliseconds. Values below 1,000 are sent as 1,000; values above 75,000 are sent as 75,000. |

#### Output

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `data` | string | Yes | Data from the action execution |
| `error` | string | No | Error if any occurred during the execution of the action |
| `successful` | boolean | Yes | Whether or not the action execution was successful or not |

### Get Agent Run

**Slug:** `SELTZ_GET_AGENT_RUN`

Retrieve one asynchronous research run to inspect or poll its pending, running, completed, failed, or cancelled state and output.

#### Input Parameters

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `run_id` | string | Yes | Agent run ID returned by Create Agent Run or List Agent Runs. |

#### Output

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `data` | string | Yes | Data from the action execution |
| `error` | string | No | Error if any occurred during the execution of the action |
| `successful` | boolean | Yes | Whether or not the action execution was successful or not |

### Get Monitor

**Slug:** `SELTZ_GET_MONITOR`

Read one live monitor's full configuration and current run state.

#### Input Parameters

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `monitor_id` | string | Yes | Monitor ID returned by Create Monitor or List Monitors. |

#### Output

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `data` | string | Yes | Data from the action execution |
| `error` | string | No | Error if any occurred during the execution of the action |
| `successful` | boolean | Yes | Whether or not the action execution was successful or not |

### Get Monitor Run

**Slug:** `SELTZ_GET_MONITOR_RUN`

Read one monitor run's status, request and record counts, timing, record bounds, and status reason.

#### Input Parameters

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `run_id` | string | Yes | Decimal-string monitor run ID. It remains a string to avoid precision loss. |
| `monitor_id` | string | Yes | Monitor that owns the run. |

#### Output

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `data` | string | Yes | Data from the action execution |
| `error` | string | No | Error if any occurred during the execution of the action |
| `successful` | boolean | Yes | Whether or not the action execution was successful or not |

### List Agent Runs

**Slug:** `SELTZ_LIST_AGENT_RUNS`

Return one newest-first page of the organization's asynchronous research runs, including status and available output.

#### Input Parameters

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `limit` | integer | No | Maximum runs to return in this page (1-100). |
| `next_cursor` | string | No | Opaque continuation cursor returned by a previous call; omit for the first page. |

#### Output

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `data` | string | Yes | Data from the action execution |
| `error` | string | No | Error if any occurred during the execution of the action |
| `successful` | boolean | Yes | Whether or not the action execution was successful or not |

### List Monitor Records

**Slug:** `SELTZ_LIST_MONITOR_RECORDS`

Return one oldest-first page of a monitor's deduplicated records, optionally restricted to records produced by one run.

#### Input Parameters

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `limit` | integer | No | Maximum records to return in this page (1-1,000); the provider's byte budget may produce a shorter page. |
| `before` | string | No | Optional exclusive upper record ID bound for a bounded window. |
| `run_id` | string | No | Optional decimal-string monitor run ID. Provide it to restrict records to one run; it remains a string to avoid precision loss. |
| `monitor_id` | string | Yes | Monitor whose records to list. |
| `next_cursor` | string | No | Continuation cursor returned by a previous call; omit for the first oldest-first page. |
| `include_content` | boolean | No | Include each record's document content and structured properties. |

#### Output

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `data` | string | Yes | Data from the action execution |
| `error` | string | No | Error if any occurred during the execution of the action |
| `successful` | boolean | Yes | Whether or not the action execution was successful or not |

### List Monitor Run Requests

**Slug:** `SELTZ_LIST_MONITOR_RUN_REQUESTS`

List every per-search outcome for one monitor run to distinguish failed searches from successful searches that found no new records.

#### Input Parameters

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `run_id` | string | Yes | Decimal-string monitor run ID. It remains a string to avoid precision loss. |
| `monitor_id` | string | Yes | Monitor that owns the run. |

#### Output

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `data` | string | Yes | Data from the action execution |
| `error` | string | No | Error if any occurred during the execution of the action |
| `successful` | boolean | Yes | Whether or not the action execution was successful or not |

### List Monitor Runs

**Slug:** `SELTZ_LIST_MONITOR_RUNS`

Return one page of a monitor's run history with status, request counts, record counts, timing, and failure reason.

#### Input Parameters

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `sort` | string ("desc" | "asc") | No | Run order: desc returns newest first; asc returns oldest first. |
| `limit` | integer | No | Maximum runs to return in this page (1-1,000). |
| `monitor_id` | string | Yes | Monitor whose run history to list. |
| `next_cursor` | string | No | Opaque continuation cursor returned by a previous call; omit for the first page. The cursor preserves the original sort direction. |

#### Output

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `data` | string | Yes | Data from the action execution |
| `error` | string | No | Error if any occurred during the execution of the action |
| `successful` | boolean | Yes | Whether or not the action execution was successful or not |

### List Monitors

**Slug:** `SELTZ_LIST_MONITORS`

Return one newest-first page of live monitors, optionally filtered by exact name or status; deleted monitors are not visible.

#### Input Parameters

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `name` | string | No | Exact monitor name to match. |
| `limit` | integer | No | Maximum monitors to return in this page (1-1,000); a byte budget may shorten it. |
| `since` | string | No | Exclusive lower monitor ID bound for a bounded window; normally omit when paging newest-first. |
| `status` | string ("active" | "paused" | "disabled") | No | Monitor lifecycle status to include. |
| `next_cursor` | string | No | Continuation cursor returned by a previous call; omit for the first newest-first page. |

#### Output

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `data` | string | Yes | Data from the action execution |
| `error` | string | No | Error if any occurred during the execution of the action |
| `successful` | boolean | Yes | Whether or not the action execution was successful or not |

### PAID Search ($3/1,000 Base, $5/1,000 Pro)

**Slug:** `SELTZ_SEARCH`

PAID: search one selected Seltz data scope for ranked documents at $3/1,000 base or $5/1,000 pro searches; people and company matches are semantic, not guaranteed exact identities.

#### Input Parameters

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `tier` | string ("base" | "pro") | No | Paid ranking tier: base costs $3/1,000 searches; pro costs $5/1,000 and adds a ranking stage. Values are trimmed and case-normalized. Defaults to base for cost safety. |
| `query` | string | Yes | Semantic search query. For people or companies, include distinguishing employer, role, location, or other identity context. |
| `scope` | string ("web" | "news" | "wikipedia" | "people" | "companies" | "financial-reports") | No | Single data index to search. Use people or companies here rather than a separate tool; those scopes rank semantic matches, not guaranteed exact identities. |
| `fields` | object | No | Optional selection and size limits for content, snippets, and properties. |
| `to_date` | string | No | Latest publication point as an ISO date or datetime, or a provider offset such as now-7d. |
| `from_date` | string | No | Earliest publication point as an ISO date or datetime, or a provider offset such as now-7d. |
| `max_results` | integer | No | Maximum ranked documents to return. Values above 1,000 are sent as 1,000. One search is billed regardless of result count. |
| `exclude_domains` | array | No | Exclude results from these domains. |
| `include_domains` | array | No | Return results only from these domains. |

#### Output

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `data` | string | Yes | Data from the action execution |
| `error` | string | No | Error if any occurred during the execution of the action |
| `successful` | boolean | Yes | Whether or not the action execution was successful or not |
