Prompts

Die Fragen, die wir im eingestellten Tracking-Rhythmus des Kontos an KI-Modelle stellen, und die Datensätze der Ausführungen, die jeder Durchlauf erzeugt.

GET`/dimensions/prompts`

Listet die Prompts eines Projekts auf.

Parameter:

project_iderforderlich
modelnur Prompts mit mindestens einer Ausführung auf diesem Modell, unabhängig von Ergebnis oder Datum
collection_id
promptPrompt-ID
country_code
language_code
from
to
page
per_page≤100

Sowohl from als auch to filtern nach dem Erstellungsdatum des Prompts.

GET`/dimensions/prompt_executions`

Listet die Prompt-Ausführungen eines Projekts auf.

Parameter:

project_iderforderlich
model
collection_id
promptPrompt-ID
country_code
language_code
prompt_type
brand_kind
from
to
page
per_page
mention_filter
citation_filter
competitorsdie unter GET /answers beschriebene zweiachsige Matrix

GET`/dimensions/query_fan_outs`

Die Fan-out-Suchanfragen, die ein Modell bei der Beantwortung Ihrer Prompts im Tracking tatsächlich gestellt hat. Damit finden Sie die Formulierungen, nach denen die Modelle suchen, statt der Formulierungen, die Sie selbst geschrieben haben. view=query (Standard) liefert für jede Fan-out-Suchanfrage eine Zeile mit ihrer Anzahl und ihrem Anteil an allen Vorkommen; view=prompt liefert für jeden Prompt eine Zeile mit der Anzahl verschiedener Fan-out-Suchanfragen, die er ausgelöst hat. Fan-out-Suchanfragen werden vor allem von ChatGPT übermittelt; ein leeres Ergebnis bedeutet daher meist, dass die ausgewählten Modelle sie nicht bereitstellen.

Parameter:

project_iderforderlich
viewquery oder prompt
queryTeilzeichenfolge im Text der Fan-out-Suchanfrage
order
direction
model
collection_id
prompt
country_code
language_code
prompt_type
brand_kind
range
from
to
page
per_page
output

GET`/suggested_prompts`

Die Warteschlange „Vorgeschlagene Prompts“, dieselbe Liste wie im Tab „Vorgeschlagene Prompts“ der App: die Vorschläge aller Generatoren, die ältesten zuerst, standardmäßig nur ausstehende.

Parameter:

project_iderforderlich
sourcewebsite, magic, topic, reddit, paa, reputational, translate, gsc oder catalog
statuspending, accepted, rejected oder all
page
per_page≤100, Standard 50

GET`/prompt_generations/:id`

Fragen Sie einen Vorschlagsdurchlauf regelmäßig ab, bis sein Status completed (mit den gespeicherten Vorschlägen) oder failed (mit Fehlercode und Fehlermeldung) lautet. Die meisten Durchläufe sind innerhalb einer Minute abgeschlossen. Durchläufe werden 24 Stunden lang aufbewahrt, ihre Vorschläge bleiben in der Warteschlange.

Parameter:

project_iderforderlich
:iddie ID des Durchlaufs, in der URL

POST`/prompts`

Erstellt bis zu 100 Prompts pro Anfrage. Bereits vorhandene Prompts (gleicher Text sowie gleiches Land und gleiche Sprache) werden übersprungen.

Anfragetext (JSON):

project_iderforderlich
promptsArray, erforderlich, maximal 100
country_codeerforderlich
language_codeerforderlich

POST`/prompts/import`

Der CSV-Import der App als JSON. Jede Zeile enthält ihr eigenes Land, ihre eigene Sprache und ihre eigenen Tags. Fehlende Tags werden erstellt. Duplikate werden übersprungen, erhalten aber trotzdem die Tags ihrer Zeile. Würden die neuen Prompts das Tariflimit überschreiten, wird nichts importiert. Eine Zeile wird übersprungen und in row_errors zurückgegeben, wenn ihr Land oder ihre Sprache unbekannt ist, wenn ein Prompt durch sie mehr als 50 Tags hätte (Zeilen mit demselben Prompt werden zusammengezählt) oder wenn die Anfrage durch ihre Tags mehr als 500 neue Tags erstellen würde. Teammitglieder benötigen die Berechtigung „Erstellen“ für Tags, um ein Tag zu erstellen (ohne sie wird eine Zeile, die ein neues Tag braucht, übersprungen und in row_errors zurückgegeben), und die Berechtigung „Aktualisieren“ für Prompts, um einem bereits vorhandenen Prompt Tags zuzuweisen (ohne sie fügt die Zeile diesem Prompt keine Tags hinzu und wird in row_errors zurückgegeben, wenn sie eines hinzugefügt hätte). Erfordert einen API-Schlüssel mit Schreibzugriff und bei Teammitgliedern die Berechtigung „Erstellen“ für Prompts.

Anfragetext (JSON): project_id (erforderlich), rows (erforderliches Array, maximal 500) aus Zeilen mit prompt (erforderlich), country_code, language_code (Codes oder Namen, standardmäßig die des Projekts) und tags (Array von Namen).

POST`/prompt_generations`

Startet einen Vorschlagsdurchlauf mit einem der Generatoren, die auch die Assistenten unter „Prompts hinzufügen“ in der App nutzen. Die Anfrage wird mit 202 beantwortet, der Durchlauf läuft im Hintergrund. Seine Vorschläge landen in der Warteschlange „Vorgeschlagene Prompts“, und ins Tracking kommt nur, was Sie übernehmen. Erfordert einen API-Schlüssel mit Schreibzugriff.

Anfragetext (JSON): project_id (erforderlich), source (erforderlich: website, keyword, topic, reddit, paa, reputational, translate oder search_console), country_code und language_code (Standard: die Werte des Projekts). Je nach Wert von source: query (bei keyword optional, bei topic und reddit erforderlich), count (keyword und topic, 1-100, Standard 20), competitor_id (keyword), search_type (reddit: topic oder brand), keywords (paa, bis zu 5), source_country_code und source_language_code (translate), min_words und all_countries (search_console, ab Tarif Growth).

POST /suggested_prompts/accept + /reject

Übernimmt ausstehende Vorschläge als Prompts ins Tracking, jeweils mit eigenem Land und eigener Sprache und optional in einem Tag, oder lehnt sie ab, sodass sie aus der Liste der ausstehenden Vorschläge verschwinden. Würden die neuen Prompts das Prompt-Limit des Tarifs überschreiten, wird nichts übernommen (ERR_LIMIT_REACHED). Erfordert einen API-Schlüssel mit Schreibzugriff.

Anfragetext (JSON):

project_iderforderlich
idserforderliches Array, maximal 100

Beim Übernehmen kann zusätzlich ein Tag über tag_ids oder tag_names angegeben werden, dazu create_missing (boolescher Wert).

PATCH`/prompts/:id`

Legt die Suchintention (prompt_type) und/oder den Markenbezug (brand_kind) eines Prompts manuell fest, genau wie beim Bearbeiten dieser Angaben auf der Seite des Prompts. Der Prompt wird als manuell kategorisiert markiert (categorization_source wird zu manual), sodass die automatische Kategorisierung ihn nicht mehr ändert; auf der Seite des Prompts lässt sich die Kategorisierung wieder auf automatisch umstellen. Aus einem White-Label-Portal heraus ist der Endpunkt nicht verfügbar (ERR_INTEGRATION_UNAVAILABLE). Erfordert einen API-Schlüssel mit Schreibzugriff und bei Teammitgliedern die Berechtigung „Aktualisieren“ für Prompts. Auch als MCP-Tool update_prompt_categorization verfügbar.

Pfad:

:idPrompt-ID

Anfragetext (JSON): project_id (erforderlich), prompt_type (informational|navigational|commercial|transactional), brand_kind (brand|brand_other|non_brand); mindestens einer der beiden. Jedes andere Feld und jeder unbekannte Wert führen zu ERR_INVALID_PARAM.

DELETE`/prompts/:id`

Löscht einen Prompt. Der Vorgang ist endgültig: Der Prompt verschwindet sofort und gibt einen Platz im Prompt-Limit frei; anschließend löscht ein Hintergrundprozess den Verlauf des Prompts (Ausführungen, Erwähnungen, Zitierungen, Sentiment-Daten). Erfordert einen API-Schlüssel mit Schreibzugriff.

Parameter:

project_iderforderlich
:idin der URL

Für Prompts außerhalb des Projekts wird ERR_NOT_FOUND zurückgegeben.

Codebeispiele

Download: OpenAPI Swagger Postman

KI fragen: In ChatGPT öffnen In Claude öffnen Als Markdown anzeigen

Sprache

Request

curl -X GET -H "Authorization: Bearer YOUR_API_KEY" \
  "https://api.llmpulse.ai/api/v1/dimensions/prompts?project_id=PROJECT_ID"

const response = await fetch("https://api.llmpulse.ai/api/v1/dimensions/prompts?project_id=PROJECT_ID", {
  headers: { "Authorization": "Bearer YOUR_API_KEY" }
});
const data = await response.json();

import requests

response = requests.get(
    "https://api.llmpulse.ai/api/v1/dimensions/prompts?project_id=PROJECT_ID",
    headers={"Authorization": "Bearer YOUR_API_KEY"}
)
data = response.json()

require 'httparty'

response = HTTParty.get(
  "https://api.llmpulse.ai/api/v1/dimensions/prompts?project_id=PROJECT_ID",
  headers: { "Authorization" => "Bearer YOUR_API_KEY" }
)
data = response.parsed_response

package main

import "net/http"

func main() {
	req, _ := http.NewRequest("GET", "https://api.llmpulse.ai/api/v1/dimensions/prompts?project_id=PROJECT_ID", nil)
	req.Header.Set("Authorization", "Bearer YOUR_API_KEY")
	client := &http.Client{}
	resp, err := client.Do(req)
	if err != nil {
		panic(err)
	}
	defer resp.Body.Close()
}

$client = new GuzzleHttp\Client();
$response = $client->request('GET', 'https://api.llmpulse.ai/api/v1/dimensions/prompts?project_id=PROJECT_ID', [
    'headers' => ['Authorization' => 'Bearer YOUR_API_KEY']
]);
$data = json_decode($response->getBody(), true);

import java.net.URI;
import java.net.http.HttpClient;
import java.net.http.HttpRequest;
import java.net.http.HttpResponse;

public class Main {
    public static void main(String[] args) throws Exception {
        HttpClient client = HttpClient.newHttpClient();
        HttpRequest request = HttpRequest.newBuilder()
            .uri(URI.create("https://api.llmpulse.ai/api/v1/dimensions/prompts?project_id=PROJECT_ID"))
            .header("Authorization", "Bearer YOUR_API_KEY")
            .GET()
            .build();
        HttpResponse<String> response = client.send(request, HttpResponse.BodyHandlers.ofString());
    }
}

Response

{ "project_id":4, "page":1, "per_page":20, "total":132,
  "data":[ { "id":101, "prompt_text":"Best AI tools for marketing?", "collection_id":7, "collection_ids":[7, 12], "country_code":"ES", "language_code":"es", "last_executed_at":"2025-03-01T10:00:00Z" } ],
  "request_id":"..." }

Request

curl -X GET -H "Authorization: Bearer YOUR_API_KEY" \
  "https://api.llmpulse.ai/api/v1/dimensions/prompt_executions?project_id=PROJECT_ID"

const response = await fetch("https://api.llmpulse.ai/api/v1/dimensions/prompt_executions?project_id=PROJECT_ID", {
  headers: { "Authorization": "Bearer YOUR_API_KEY" }
});
const data = await response.json();

import requests

response = requests.get(
    "https://api.llmpulse.ai/api/v1/dimensions/prompt_executions?project_id=PROJECT_ID",
    headers={"Authorization": "Bearer YOUR_API_KEY"}
)
data = response.json()

require 'httparty'

response = HTTParty.get(
  "https://api.llmpulse.ai/api/v1/dimensions/prompt_executions?project_id=PROJECT_ID",
  headers: { "Authorization" => "Bearer YOUR_API_KEY" }
)
data = response.parsed_response

package main

import "net/http"

func main() {
	req, _ := http.NewRequest("GET", "https://api.llmpulse.ai/api/v1/dimensions/prompt_executions?project_id=PROJECT_ID", nil)
	req.Header.Set("Authorization", "Bearer YOUR_API_KEY")
	client := &http.Client{}
	resp, err := client.Do(req)
	if err != nil {
		panic(err)
	}
	defer resp.Body.Close()
}

$client = new GuzzleHttp\Client();
$response = $client->request('GET', 'https://api.llmpulse.ai/api/v1/dimensions/prompt_executions?project_id=PROJECT_ID', [
    'headers' => ['Authorization' => 'Bearer YOUR_API_KEY']
]);
$data = json_decode($response->getBody(), true);

import java.net.URI;
import java.net.http.HttpClient;
import java.net.http.HttpRequest;
import java.net.http.HttpResponse;

public class Main {
    public static void main(String[] args) throws Exception {
        HttpClient client = HttpClient.newHttpClient();
        HttpRequest request = HttpRequest.newBuilder()
            .uri(URI.create("https://api.llmpulse.ai/api/v1/dimensions/prompt_executions?project_id=PROJECT_ID"))
            .header("Authorization", "Bearer YOUR_API_KEY")
            .GET()
            .build();
        HttpResponse<String> response = client.send(request, HttpResponse.BodyHandlers.ofString());
    }
}

Response

{ "project_id":4, "page":1, "per_page":20, "total":540,
  "data":[ { "id":9001, "prompt_id":101, "executed_at":"2025-03-01T10:00:00Z", "duration":1.23, "success":true, "model":"gpt-4o" } ],
  "request_id":"..." }

Request

curl -X GET -H "Authorization: Bearer YOUR_API_KEY" \
  "https://api.llmpulse.ai/api/v1/suggested_prompts?project_id=PROJECT_ID&status=pending"

const response = await fetch("https://api.llmpulse.ai/api/v1/suggested_prompts?project_id=PROJECT_ID&status=pending", {
  headers: { "Authorization": "Bearer YOUR_API_KEY" }
});
const data = await response.json();

import requests

response = requests.get(
    "https://api.llmpulse.ai/api/v1/suggested_prompts?project_id=PROJECT_ID&status=pending",
    headers={"Authorization": "Bearer YOUR_API_KEY"}
)
data = response.json()

require 'httparty'

response = HTTParty.get(
  "https://api.llmpulse.ai/api/v1/suggested_prompts?project_id=PROJECT_ID&status=pending",
  headers: { "Authorization" => "Bearer YOUR_API_KEY" }
)
data = response.parsed_response

package main

import "net/http"

func main() {
	req, _ := http.NewRequest("GET", "https://api.llmpulse.ai/api/v1/suggested_prompts?project_id=PROJECT_ID&status=pending", nil)
	req.Header.Set("Authorization", "Bearer YOUR_API_KEY")
	client := &http.Client{}
	resp, err := client.Do(req)
	if err != nil {
		panic(err)
	}
	defer resp.Body.Close()
}

$client = new GuzzleHttp\Client();
$response = $client->request('GET', 'https://api.llmpulse.ai/api/v1/suggested_prompts?project_id=PROJECT_ID&status=pending', [
    'headers' => ['Authorization' => 'Bearer YOUR_API_KEY']
]);
$data = json_decode($response->getBody(), true);

import java.net.URI;
import java.net.http.HttpClient;
import java.net.http.HttpRequest;
import java.net.http.HttpResponse;

public class Main {
    public static void main(String[] args) throws Exception {
        HttpClient client = HttpClient.newHttpClient();
        HttpRequest request = HttpRequest.newBuilder()
            .uri(URI.create("https://api.llmpulse.ai/api/v1/suggested_prompts?project_id=PROJECT_ID&status=pending"))
            .header("Authorization", "Bearer YOUR_API_KEY")
            .GET()
            .build();
        HttpResponse<String> response = client.send(request, HttpResponse.BodyHandlers.ofString());
    }
}

Response

{ "project_id":123, "page":1, "per_page":50, "total":18,
  "data":[ { "id":77, "prompt":"best running shoes for flat feet", "source":"magic", "status":"pending", "country_code":"US", "language_code":"en", "prompt_id":null, "accepted_at":null, "created_at":"2026-10-15T09:01:12Z" } ],
  "request_id":"..." }

Request

curl -X GET -H "Authorization: Bearer YOUR_API_KEY" \
  "https://api.llmpulse.ai/api/v1/prompt_generations/6f1c2a7e-0d4b-4c39-9a51-2f7b8e3d1c90?project_id=PROJECT_ID"

const response = await fetch("https://api.llmpulse.ai/api/v1/prompt_generations/6f1c2a7e-0d4b-4c39-9a51-2f7b8e3d1c90?project_id=PROJECT_ID", {
  headers: { "Authorization": "Bearer YOUR_API_KEY" }
});
const data = await response.json();

import requests

response = requests.get(
    "https://api.llmpulse.ai/api/v1/prompt_generations/6f1c2a7e-0d4b-4c39-9a51-2f7b8e3d1c90?project_id=PROJECT_ID",
    headers={"Authorization": "Bearer YOUR_API_KEY"}
)
data = response.json()

require 'httparty'

response = HTTParty.get(
  "https://api.llmpulse.ai/api/v1/prompt_generations/6f1c2a7e-0d4b-4c39-9a51-2f7b8e3d1c90?project_id=PROJECT_ID",
  headers: { "Authorization" => "Bearer YOUR_API_KEY" }
)
data = response.parsed_response

package main

import "net/http"

func main() {
	req, _ := http.NewRequest("GET", "https://api.llmpulse.ai/api/v1/prompt_generations/6f1c2a7e-0d4b-4c39-9a51-2f7b8e3d1c90?project_id=PROJECT_ID", nil)
	req.Header.Set("Authorization", "Bearer YOUR_API_KEY")
	client := &http.Client{}
	resp, err := client.Do(req)
	if err != nil {
		panic(err)
	}
	defer resp.Body.Close()
}

$client = new GuzzleHttp\Client();
$response = $client->request('GET', 'https://api.llmpulse.ai/api/v1/prompt_generations/6f1c2a7e-0d4b-4c39-9a51-2f7b8e3d1c90?project_id=PROJECT_ID', [
    'headers' => ['Authorization' => 'Bearer YOUR_API_KEY']
]);
$data = json_decode($response->getBody(), true);

import java.net.URI;
import java.net.http.HttpClient;
import java.net.http.HttpRequest;
import java.net.http.HttpResponse;

public class Main {
    public static void main(String[] args) throws Exception {
        HttpClient client = HttpClient.newHttpClient();
        HttpRequest request = HttpRequest.newBuilder()
            .uri(URI.create("https://api.llmpulse.ai/api/v1/prompt_generations/6f1c2a7e-0d4b-4c39-9a51-2f7b8e3d1c90?project_id=PROJECT_ID"))
            .header("Authorization", "Bearer YOUR_API_KEY")
            .GET()
            .build();
        HttpResponse<String> response = client.send(request, HttpResponse.BodyHandlers.ofString());
    }
}

Request

curl -X POST -H "Authorization: Bearer YOUR_API_KEY" -H "Content-Type: application/json" -d '{"project_id":"PROJECT_ID","prompts":["best CRM software 2026","top project management tools"],"country_code":"US","language_code":"en"}' \
  "https://api.llmpulse.ai/api/v1/prompts"

const response = await fetch("https://api.llmpulse.ai/api/v1/prompts", {
  method: "POST",
  headers: {
    "Authorization": "Bearer YOUR_API_KEY",
    "Content-Type": "application/json"
  },
  body: JSON.stringify({"project_id":"PROJECT_ID","prompts":["best CRM software 2026","top project management tools"],"country_code":"US","language_code":"en"})
});
const data = await response.json();

import requests

response = requests.post(
    "https://api.llmpulse.ai/api/v1/prompts",
    headers={"Authorization": "Bearer YOUR_API_KEY"},
    json={"project_id": "PROJECT_ID", "prompts": ["best CRM software 2026", "top project management tools"], "country_code": "US", "language_code": "en"}
)
data = response.json()

require 'httparty'
require 'json'

response = HTTParty.post(
  "https://api.llmpulse.ai/api/v1/prompts",
  headers: {
    "Authorization" => "Bearer YOUR_API_KEY",
    "Content-Type" => "application/json"
  },
  body: { "project_id" => "PROJECT_ID", "prompts" => ["best CRM software 2026", "top project management tools"], "country_code" => "US", "language_code" => "en" }.to_json
)
data = response.parsed_response

package main

import (
	"net/http"
	"strings"
)

func main() {
	req, _ := http.NewRequest("POST", "https://api.llmpulse.ai/api/v1/prompts", strings.NewReader(`{"project_id":"PROJECT_ID","prompts":["best CRM software 2026","top project management tools"],"country_code":"US","language_code":"en"}`))
	req.Header.Set("Authorization", "Bearer YOUR_API_KEY")
	req.Header.Set("Content-Type", "application/json")
	client := &http.Client{}
	resp, err := client.Do(req)
	if err != nil {
		panic(err)
	}
	defer resp.Body.Close()
}

$client = new GuzzleHttp\Client();
$response = $client->request('POST', 'https://api.llmpulse.ai/api/v1/prompts', [
    'headers' => ['Authorization' => 'Bearer YOUR_API_KEY'],
    'json' => ['project_id' => 'PROJECT_ID', 'prompts' => ['best CRM software 2026', 'top project management tools'], 'country_code' => 'US', 'language_code' => 'en']
]);
$data = json_decode($response->getBody(), true);

import java.net.URI;
import java.net.http.HttpClient;
import java.net.http.HttpRequest;
import java.net.http.HttpResponse;

public class Main {
    public static void main(String[] args) throws Exception {
        HttpClient client = HttpClient.newHttpClient();
        HttpRequest request = HttpRequest.newBuilder()
            .uri(URI.create("https://api.llmpulse.ai/api/v1/prompts"))
            .header("Authorization", "Bearer YOUR_API_KEY")
            .header("Content-Type", "application/json")
            .POST(HttpRequest.BodyPublishers.ofString("{\"project_id\":\"PROJECT_ID\",\"prompts\":[\"best CRM software 2026\",\"top project management tools\"],\"country_code\":\"US\",\"language_code\":\"en\"}"))
            .build();
        HttpResponse<String> response = client.send(request, HttpResponse.BodyHandlers.ofString());
    }
}

Response

{
  "project_id": 123,
  "created": 2,
  "skipped": 1,
  "total_after": 53,
  "prompts_available": 47,
  "data": [
    { "id": 456, "raw_input": "best CRM software 2026", "status": "created" },
    { "id": 457, "raw_input": "top project management tools", "status": "created" },
    { "id": 400, "raw_input": "best AI tools", "status": "exists" }
  ],
  "request_id": "..."
}

Request

curl -X POST -H "Authorization: Bearer YOUR_API_KEY" -H "Content-Type: application/json" -d '{"project_id":"PROJECT_ID","rows":[{"prompt":"best CRM for startups","country_code":"US","language_code":"en","tags":["CRM"]},{"prompt":"mejor CRM para pymes","country_code":"ES","language_code":"es"}]}' \
  "https://api.llmpulse.ai/api/v1/prompts/import"

const response = await fetch("https://api.llmpulse.ai/api/v1/prompts/import", {
  method: "POST",
  headers: {
    "Authorization": "Bearer YOUR_API_KEY",
    "Content-Type": "application/json"
  },
  body: JSON.stringify({"project_id":"PROJECT_ID","rows":[{"prompt":"best CRM for startups","country_code":"US","language_code":"en","tags":["CRM"]},{"prompt":"mejor CRM para pymes","country_code":"ES","language_code":"es"}]})
});
const data = await response.json();

import requests

response = requests.post(
    "https://api.llmpulse.ai/api/v1/prompts/import",
    headers={"Authorization": "Bearer YOUR_API_KEY"},
    json={"project_id": "PROJECT_ID", "rows": [{"prompt": "best CRM for startups", "country_code": "US", "language_code": "en", "tags": ["CRM"]}, {"prompt": "mejor CRM para pymes", "country_code": "ES", "language_code": "es"}]}
)
data = response.json()

require 'httparty'
require 'json'

response = HTTParty.post(
  "https://api.llmpulse.ai/api/v1/prompts/import",
  headers: {
    "Authorization" => "Bearer YOUR_API_KEY",
    "Content-Type" => "application/json"
  },
  body: { "project_id" => "PROJECT_ID", "rows" => [{ "prompt" => "best CRM for startups", "country_code" => "US", "language_code" => "en", "tags" => ["CRM"] }, { "prompt" => "mejor CRM para pymes", "country_code" => "ES", "language_code" => "es" }] }.to_json
)
data = response.parsed_response

package main

import (
	"net/http"
	"strings"
)

func main() {
	req, _ := http.NewRequest("POST", "https://api.llmpulse.ai/api/v1/prompts/import", strings.NewReader(`{"project_id":"PROJECT_ID","rows":[{"prompt":"best CRM for startups","country_code":"US","language_code":"en","tags":["CRM"]},{"prompt":"mejor CRM para pymes","country_code":"ES","language_code":"es"}]}`))
	req.Header.Set("Authorization", "Bearer YOUR_API_KEY")
	req.Header.Set("Content-Type", "application/json")
	client := &http.Client{}
	resp, err := client.Do(req)
	if err != nil {
		panic(err)
	}
	defer resp.Body.Close()
}

$client = new GuzzleHttp\Client();
$response = $client->request('POST', 'https://api.llmpulse.ai/api/v1/prompts/import', [
    'headers' => ['Authorization' => 'Bearer YOUR_API_KEY'],
    'json' => ['project_id' => 'PROJECT_ID', 'rows' => [['prompt' => 'best CRM for startups', 'country_code' => 'US', 'language_code' => 'en', 'tags' => ['CRM']], ['prompt' => 'mejor CRM para pymes', 'country_code' => 'ES', 'language_code' => 'es']]]
]);
$data = json_decode($response->getBody(), true);

import java.net.URI;
import java.net.http.HttpClient;
import java.net.http.HttpRequest;
import java.net.http.HttpResponse;

public class Main {
    public static void main(String[] args) throws Exception {
        HttpClient client = HttpClient.newHttpClient();
        HttpRequest request = HttpRequest.newBuilder()
            .uri(URI.create("https://api.llmpulse.ai/api/v1/prompts/import"))
            .header("Authorization", "Bearer YOUR_API_KEY")
            .header("Content-Type", "application/json")
            .POST(HttpRequest.BodyPublishers.ofString("{\"project_id\":\"PROJECT_ID\",\"rows\":[{\"prompt\":\"best CRM for startups\",\"country_code\":\"US\",\"language_code\":\"en\",\"tags\":[\"CRM\"]},{\"prompt\":\"mejor CRM para pymes\",\"country_code\":\"ES\",\"language_code\":\"es\"}]}"))
            .build();
        HttpResponse<String> response = client.send(request, HttpResponse.BodyHandlers.ofString());
    }
}

Response

{
  "project_id": 123,
  "processed_rows": 2,
  "created": 2,
  "duplicates": 0,
  "tagged": 2,
  "created_tags": 1,
  "row_errors": [],
  "length_warning": null,
  "prompts_available": 46,
  "request_id": "..."
}

Request

curl -X POST -H "Authorization: Bearer YOUR_API_KEY" -H "Content-Type: application/json" -d '{"project_id":"PROJECT_ID","source":"topic","query":"electric SUVs","count":20}' \
  "https://api.llmpulse.ai/api/v1/prompt_generations"

const response = await fetch("https://api.llmpulse.ai/api/v1/prompt_generations", {
  method: "POST",
  headers: {
    "Authorization": "Bearer YOUR_API_KEY",
    "Content-Type": "application/json"
  },
  body: JSON.stringify({"project_id":"PROJECT_ID","source":"topic","query":"electric SUVs","count":20})
});
const data = await response.json();

import requests

response = requests.post(
    "https://api.llmpulse.ai/api/v1/prompt_generations",
    headers={"Authorization": "Bearer YOUR_API_KEY"},
    json={"project_id": "PROJECT_ID", "source": "topic", "query": "electric SUVs", "count": 20}
)
data = response.json()

require 'httparty'
require 'json'

response = HTTParty.post(
  "https://api.llmpulse.ai/api/v1/prompt_generations",
  headers: {
    "Authorization" => "Bearer YOUR_API_KEY",
    "Content-Type" => "application/json"
  },
  body: { "project_id" => "PROJECT_ID", "source" => "topic", "query" => "electric SUVs", "count" => 20 }.to_json
)
data = response.parsed_response

package main

import (
	"net/http"
	"strings"
)

func main() {
	req, _ := http.NewRequest("POST", "https://api.llmpulse.ai/api/v1/prompt_generations", strings.NewReader(`{"project_id":"PROJECT_ID","source":"topic","query":"electric SUVs","count":20}`))
	req.Header.Set("Authorization", "Bearer YOUR_API_KEY")
	req.Header.Set("Content-Type", "application/json")
	client := &http.Client{}
	resp, err := client.Do(req)
	if err != nil {
		panic(err)
	}
	defer resp.Body.Close()
}

$client = new GuzzleHttp\Client();
$response = $client->request('POST', 'https://api.llmpulse.ai/api/v1/prompt_generations', [
    'headers' => ['Authorization' => 'Bearer YOUR_API_KEY'],
    'json' => ['project_id' => 'PROJECT_ID', 'source' => 'topic', 'query' => 'electric SUVs', 'count' => 20]
]);
$data = json_decode($response->getBody(), true);

import java.net.URI;
import java.net.http.HttpClient;
import java.net.http.HttpRequest;
import java.net.http.HttpResponse;

public class Main {
    public static void main(String[] args) throws Exception {
        HttpClient client = HttpClient.newHttpClient();
        HttpRequest request = HttpRequest.newBuilder()
            .uri(URI.create("https://api.llmpulse.ai/api/v1/prompt_generations"))
            .header("Authorization", "Bearer YOUR_API_KEY")
            .header("Content-Type", "application/json")
            .POST(HttpRequest.BodyPublishers.ofString("{\"project_id\":\"PROJECT_ID\",\"source\":\"topic\",\"query\":\"electric SUVs\",\"count\":20}"))
            .build();
        HttpResponse<String> response = client.send(request, HttpResponse.BodyHandlers.ofString());
    }
}

Response

{
  "id": "6f1c2a7e-0d4b-4c39-9a51-2f7b8e3d1c90",
  "project_id": 123,
  "source": "topic",
  "status": "processing",
  "params": { "country_code": "US", "language_code": "en", "query": "electric SUVs", "count": 20 },
  "created_at": "2026-10-15T09:00:00Z",
  "finished_at": null,
  "suggestions_count": 0,
  "suggestions": [],
  "error": null,
  "request_id": "..."
}

Request

curl -X POST -H "Authorization: Bearer YOUR_API_KEY" -H "Content-Type: application/json" -d '{"project_id":"PROJECT_ID","ids":[77,78],"tag_names":["Electric"],"create_missing":true}' \
  "https://api.llmpulse.ai/api/v1/suggested_prompts/accept"

const response = await fetch("https://api.llmpulse.ai/api/v1/suggested_prompts/accept", {
  method: "POST",
  headers: {
    "Authorization": "Bearer YOUR_API_KEY",
    "Content-Type": "application/json"
  },
  body: JSON.stringify({"project_id":"PROJECT_ID","ids":[77,78],"tag_names":["Electric"],"create_missing":true})
});
const data = await response.json();

import requests

response = requests.post(
    "https://api.llmpulse.ai/api/v1/suggested_prompts/accept",
    headers={"Authorization": "Bearer YOUR_API_KEY"},
    json={"project_id": "PROJECT_ID", "ids": [77, 78], "tag_names": ["Electric"], "create_missing": True}
)
data = response.json()

require 'httparty'
require 'json'

response = HTTParty.post(
  "https://api.llmpulse.ai/api/v1/suggested_prompts/accept",
  headers: {
    "Authorization" => "Bearer YOUR_API_KEY",
    "Content-Type" => "application/json"
  },
  body: { "project_id" => "PROJECT_ID", "ids" => [77, 78], "tag_names" => ["Electric"], "create_missing" => true }.to_json
)
data = response.parsed_response

package main

import (
	"net/http"
	"strings"
)

func main() {
	req, _ := http.NewRequest("POST", "https://api.llmpulse.ai/api/v1/suggested_prompts/accept", strings.NewReader(`{"project_id":"PROJECT_ID","ids":[77,78],"tag_names":["Electric"],"create_missing":true}`))
	req.Header.Set("Authorization", "Bearer YOUR_API_KEY")
	req.Header.Set("Content-Type", "application/json")
	client := &http.Client{}
	resp, err := client.Do(req)
	if err != nil {
		panic(err)
	}
	defer resp.Body.Close()
}

$client = new GuzzleHttp\Client();
$response = $client->request('POST', 'https://api.llmpulse.ai/api/v1/suggested_prompts/accept', [
    'headers' => ['Authorization' => 'Bearer YOUR_API_KEY'],
    'json' => ['project_id' => 'PROJECT_ID', 'ids' => [77, 78], 'tag_names' => ['Electric'], 'create_missing' => true]
]);
$data = json_decode($response->getBody(), true);

import java.net.URI;
import java.net.http.HttpClient;
import java.net.http.HttpRequest;
import java.net.http.HttpResponse;

public class Main {
    public static void main(String[] args) throws Exception {
        HttpClient client = HttpClient.newHttpClient();
        HttpRequest request = HttpRequest.newBuilder()
            .uri(URI.create("https://api.llmpulse.ai/api/v1/suggested_prompts/accept"))
            .header("Authorization", "Bearer YOUR_API_KEY")
            .header("Content-Type", "application/json")
            .POST(HttpRequest.BodyPublishers.ofString("{\"project_id\":\"PROJECT_ID\",\"ids\":[77,78],\"tag_names\":[\"Electric\"],\"create_missing\":true}"))
            .build();
        HttpResponse<String> response = client.send(request, HttpResponse.BodyHandlers.ofString());
    }
}

Response

{
  "project_id": 123,
  "accepted": [ { "suggestion_id": 77, "prompt_id": 501, "collection_id": 12 } ],
  "skipped": [ { "suggestion_id": 78, "reason": "rejected" } ],
  "prompts_available": 44,
  "request_id": "..."
}

Request

curl -X PATCH -H "Authorization: Bearer YOUR_API_KEY" -H "Content-Type: application/json" -d '{"project_id":"PROJECT_ID","prompt_type":"commercial","brand_kind":"non_brand"}' \
  "https://api.llmpulse.ai/api/v1/prompts/123"

const response = await fetch("https://api.llmpulse.ai/api/v1/prompts/123", {
  method: "PATCH",
  headers: {
    "Authorization": "Bearer YOUR_API_KEY",
    "Content-Type": "application/json"
  },
  body: JSON.stringify({"project_id":"PROJECT_ID","prompt_type":"commercial","brand_kind":"non_brand"})
});
const data = await response.json();

import requests

response = requests.patch(
    "https://api.llmpulse.ai/api/v1/prompts/123",
    headers={"Authorization": "Bearer YOUR_API_KEY"},
    json={"project_id": "PROJECT_ID", "prompt_type": "commercial", "brand_kind": "non_brand"}
)
data = response.json()

require 'httparty'
require 'json'

response = HTTParty.patch(
  "https://api.llmpulse.ai/api/v1/prompts/123",
  headers: {
    "Authorization" => "Bearer YOUR_API_KEY",
    "Content-Type" => "application/json"
  },
  body: { "project_id" => "PROJECT_ID", "prompt_type" => "commercial", "brand_kind" => "non_brand" }.to_json
)
data = response.parsed_response

package main

import (
	"net/http"
	"strings"
)

func main() {
	req, _ := http.NewRequest("PATCH", "https://api.llmpulse.ai/api/v1/prompts/123", strings.NewReader(`{"project_id":"PROJECT_ID","prompt_type":"commercial","brand_kind":"non_brand"}`))
	req.Header.Set("Authorization", "Bearer YOUR_API_KEY")
	req.Header.Set("Content-Type", "application/json")
	client := &http.Client{}
	resp, err := client.Do(req)
	if err != nil {
		panic(err)
	}
	defer resp.Body.Close()
}

$client = new GuzzleHttp\Client();
$response = $client->request('PATCH', 'https://api.llmpulse.ai/api/v1/prompts/123', [
    'headers' => ['Authorization' => 'Bearer YOUR_API_KEY'],
    'json' => ['project_id' => 'PROJECT_ID', 'prompt_type' => 'commercial', 'brand_kind' => 'non_brand']
]);
$data = json_decode($response->getBody(), true);

import java.net.URI;
import java.net.http.HttpClient;
import java.net.http.HttpRequest;
import java.net.http.HttpResponse;

public class Main {
    public static void main(String[] args) throws Exception {
        HttpClient client = HttpClient.newHttpClient();
        HttpRequest request = HttpRequest.newBuilder()
            .uri(URI.create("https://api.llmpulse.ai/api/v1/prompts/123"))
            .header("Authorization", "Bearer YOUR_API_KEY")
            .header("Content-Type", "application/json")
            .method("PATCH", HttpRequest.BodyPublishers.ofString("{\"project_id\":\"PROJECT_ID\",\"prompt_type\":\"commercial\",\"brand_kind\":\"non_brand\"}"))
            .build();
        HttpResponse<String> response = client.send(request, HttpResponse.BodyHandlers.ofString());
    }
}

Prompts

GET/dimensions/prompts

GET/dimensions/prompt_executions

GET/dimensions/query_fan_outs

GET/suggested_prompts

GET/prompt_generations/:id

POST/prompts

POST/prompts/import

POST/prompt_generations

POST /suggested_prompts/accept + /reject

PATCH/prompts/:id

DELETE/prompts/:id

GET`/dimensions/prompts`

GET`/dimensions/prompt_executions`

GET`/dimensions/query_fan_outs`

GET`/suggested_prompts`

GET`/prompt_generations/:id`

POST`/prompts`

POST`/prompts/import`

POST`/prompt_generations`

PATCH`/prompts/:id`

DELETE`/prompts/:id`