curl --request POST \
--url https://api.globalwebindex.com/v1/spark-api/mcp \
--header 'Authorization: <api-key>' \
--header 'Content-Type: application/json' \
--data '
{
"jsonrpc": "2.0",
"id": "req-123",
"method": "tools/call",
"params": {
"name": "chat_gwi",
"arguments": {
"prompt": "What marketing channels work best for Audi drivers in the US?"
}
}
}
'import requests
url = "https://api.globalwebindex.com/v1/spark-api/mcp"
payload = {
"jsonrpc": "2.0",
"id": "req-123",
"method": "tools/call",
"params": {
"name": "chat_gwi",
"arguments": { "prompt": "What marketing channels work best for Audi drivers in the US?" }
}
}
headers = {
"Authorization": "<api-key>",
"Content-Type": "application/json"
}
response = requests.post(url, json=payload, headers=headers)
print(response.text)const options = {
method: 'POST',
headers: {Authorization: '<api-key>', 'Content-Type': 'application/json'},
body: JSON.stringify({
jsonrpc: '2.0',
id: 'req-123',
method: 'tools/call',
params: {
name: 'chat_gwi',
arguments: {prompt: 'What marketing channels work best for Audi drivers in the US?'}
}
})
};
fetch('https://api.globalwebindex.com/v1/spark-api/mcp', options)
.then(res => res.json())
.then(res => console.log(res))
.catch(err => console.error(err));<?php
$curl = curl_init();
curl_setopt_array($curl, [
CURLOPT_URL => "https://api.globalwebindex.com/v1/spark-api/mcp",
CURLOPT_RETURNTRANSFER => true,
CURLOPT_ENCODING => "",
CURLOPT_MAXREDIRS => 10,
CURLOPT_TIMEOUT => 30,
CURLOPT_HTTP_VERSION => CURL_HTTP_VERSION_1_1,
CURLOPT_CUSTOMREQUEST => "POST",
CURLOPT_POSTFIELDS => json_encode([
'jsonrpc' => '2.0',
'id' => 'req-123',
'method' => 'tools/call',
'params' => [
'name' => 'chat_gwi',
'arguments' => [
'prompt' => 'What marketing channels work best for Audi drivers in the US?'
]
]
]),
CURLOPT_HTTPHEADER => [
"Authorization: <api-key>",
"Content-Type: application/json"
],
]);
$response = curl_exec($curl);
$err = curl_error($curl);
curl_close($curl);
if ($err) {
echo "cURL Error #:" . $err;
} else {
echo $response;
}package main
import (
"fmt"
"strings"
"net/http"
"io"
)
func main() {
url := "https://api.globalwebindex.com/v1/spark-api/mcp"
payload := strings.NewReader("{\n \"jsonrpc\": \"2.0\",\n \"id\": \"req-123\",\n \"method\": \"tools/call\",\n \"params\": {\n \"name\": \"chat_gwi\",\n \"arguments\": {\n \"prompt\": \"What marketing channels work best for Audi drivers in the US?\"\n }\n }\n}")
req, _ := http.NewRequest("POST", url, payload)
req.Header.Add("Authorization", "<api-key>")
req.Header.Add("Content-Type", "application/json")
res, _ := http.DefaultClient.Do(req)
defer res.Body.Close()
body, _ := io.ReadAll(res.Body)
fmt.Println(string(body))
}HttpResponse<String> response = Unirest.post("https://api.globalwebindex.com/v1/spark-api/mcp")
.header("Authorization", "<api-key>")
.header("Content-Type", "application/json")
.body("{\n \"jsonrpc\": \"2.0\",\n \"id\": \"req-123\",\n \"method\": \"tools/call\",\n \"params\": {\n \"name\": \"chat_gwi\",\n \"arguments\": {\n \"prompt\": \"What marketing channels work best for Audi drivers in the US?\"\n }\n }\n}")
.asString();require 'uri'
require 'net/http'
url = URI("https://api.globalwebindex.com/v1/spark-api/mcp")
http = Net::HTTP.new(url.host, url.port)
http.use_ssl = true
request = Net::HTTP::Post.new(url)
request["Authorization"] = '<api-key>'
request["Content-Type"] = 'application/json'
request.body = "{\n \"jsonrpc\": \"2.0\",\n \"id\": \"req-123\",\n \"method\": \"tools/call\",\n \"params\": {\n \"name\": \"chat_gwi\",\n \"arguments\": {\n \"prompt\": \"What marketing channels work best for Audi drivers in the US?\"\n }\n }\n}"
response = http.request(request)
puts response.read_body{
"jsonrpc": "2.0",
"id": "<string>",
"result": {
"content": [
{
"type": "text",
"text": "<string>"
}
],
"isError": true,
"_meta": {
"isInternalError": true
}
}
}{
"error": "unauthorized",
"message": "Invalid or missing bearer token",
"code": 401
}Execute GWI MCP tool calls
This endpoint handles MCP tool calls through JSON-RPC. Supported tools:
1. chat_gwi - Query audience insights
- Returns main response, insights with IDs, sources, and chat ID
- Use for initial queries and follow-ups
2. explore_insight_gwi - Get detailed statistics
- Returns percentages, sample sizes, and index scores
- Use with insight IDs from chat_gwi responses
3. list_datasets - List available datasets
- Used to get the dataset codes that can be passed to chat_gwi
- Returns hierarchical folder and dataset nodes so clients can identify valid root/child combinations
4. search_audiences - Search available audiences
- Returns matching audience entries with IDs, metadata, and dataset availability
- Supports optional dataset filtering, exclusion IDs, and result limits
curl --request POST \
--url https://api.globalwebindex.com/v1/spark-api/mcp \
--header 'Authorization: <api-key>' \
--header 'Content-Type: application/json' \
--data '
{
"jsonrpc": "2.0",
"id": "req-123",
"method": "tools/call",
"params": {
"name": "chat_gwi",
"arguments": {
"prompt": "What marketing channels work best for Audi drivers in the US?"
}
}
}
'import requests
url = "https://api.globalwebindex.com/v1/spark-api/mcp"
payload = {
"jsonrpc": "2.0",
"id": "req-123",
"method": "tools/call",
"params": {
"name": "chat_gwi",
"arguments": { "prompt": "What marketing channels work best for Audi drivers in the US?" }
}
}
headers = {
"Authorization": "<api-key>",
"Content-Type": "application/json"
}
response = requests.post(url, json=payload, headers=headers)
print(response.text)const options = {
method: 'POST',
headers: {Authorization: '<api-key>', 'Content-Type': 'application/json'},
body: JSON.stringify({
jsonrpc: '2.0',
id: 'req-123',
method: 'tools/call',
params: {
name: 'chat_gwi',
arguments: {prompt: 'What marketing channels work best for Audi drivers in the US?'}
}
})
};
fetch('https://api.globalwebindex.com/v1/spark-api/mcp', options)
.then(res => res.json())
.then(res => console.log(res))
.catch(err => console.error(err));<?php
$curl = curl_init();
curl_setopt_array($curl, [
CURLOPT_URL => "https://api.globalwebindex.com/v1/spark-api/mcp",
CURLOPT_RETURNTRANSFER => true,
CURLOPT_ENCODING => "",
CURLOPT_MAXREDIRS => 10,
CURLOPT_TIMEOUT => 30,
CURLOPT_HTTP_VERSION => CURL_HTTP_VERSION_1_1,
CURLOPT_CUSTOMREQUEST => "POST",
CURLOPT_POSTFIELDS => json_encode([
'jsonrpc' => '2.0',
'id' => 'req-123',
'method' => 'tools/call',
'params' => [
'name' => 'chat_gwi',
'arguments' => [
'prompt' => 'What marketing channels work best for Audi drivers in the US?'
]
]
]),
CURLOPT_HTTPHEADER => [
"Authorization: <api-key>",
"Content-Type: application/json"
],
]);
$response = curl_exec($curl);
$err = curl_error($curl);
curl_close($curl);
if ($err) {
echo "cURL Error #:" . $err;
} else {
echo $response;
}package main
import (
"fmt"
"strings"
"net/http"
"io"
)
func main() {
url := "https://api.globalwebindex.com/v1/spark-api/mcp"
payload := strings.NewReader("{\n \"jsonrpc\": \"2.0\",\n \"id\": \"req-123\",\n \"method\": \"tools/call\",\n \"params\": {\n \"name\": \"chat_gwi\",\n \"arguments\": {\n \"prompt\": \"What marketing channels work best for Audi drivers in the US?\"\n }\n }\n}")
req, _ := http.NewRequest("POST", url, payload)
req.Header.Add("Authorization", "<api-key>")
req.Header.Add("Content-Type", "application/json")
res, _ := http.DefaultClient.Do(req)
defer res.Body.Close()
body, _ := io.ReadAll(res.Body)
fmt.Println(string(body))
}HttpResponse<String> response = Unirest.post("https://api.globalwebindex.com/v1/spark-api/mcp")
.header("Authorization", "<api-key>")
.header("Content-Type", "application/json")
.body("{\n \"jsonrpc\": \"2.0\",\n \"id\": \"req-123\",\n \"method\": \"tools/call\",\n \"params\": {\n \"name\": \"chat_gwi\",\n \"arguments\": {\n \"prompt\": \"What marketing channels work best for Audi drivers in the US?\"\n }\n }\n}")
.asString();require 'uri'
require 'net/http'
url = URI("https://api.globalwebindex.com/v1/spark-api/mcp")
http = Net::HTTP.new(url.host, url.port)
http.use_ssl = true
request = Net::HTTP::Post.new(url)
request["Authorization"] = '<api-key>'
request["Content-Type"] = 'application/json'
request.body = "{\n \"jsonrpc\": \"2.0\",\n \"id\": \"req-123\",\n \"method\": \"tools/call\",\n \"params\": {\n \"name\": \"chat_gwi\",\n \"arguments\": {\n \"prompt\": \"What marketing channels work best for Audi drivers in the US?\"\n }\n }\n}"
response = http.request(request)
puts response.read_body{
"jsonrpc": "2.0",
"id": "<string>",
"result": {
"content": [
{
"type": "text",
"text": "<string>"
}
],
"isError": true,
"_meta": {
"isInternalError": true
}
}
}{
"error": "unauthorized",
"message": "Invalid or missing bearer token",
"code": 401
}Authorizations
Format: Bearer YOUR_TOKEN
Body
JSON-RPC request for any supported MCP tool
- chat_gwi
- explore_insight_gwi
- list_datasets
- search_audiences
Query audience insights:
- Returns main response, insights with IDs, sources, and chat ID
- Use for initial queries and follow-ups
Response
JSON-RPC response. HTTP 200 means the JSON-RPC call was processed, not that the tool succeeded.
Inspect result.isError if present:
true— tool failed; the message is inresult.content[0].text, andresult._meta.isInternalErrormay betruefor internal or timeout-style failures.
See Tool errors (result.isError) in the overview above for full behavior.
JSON-RPC 2.0 success envelope for MCP tools/call (no top-level error field).
When the tool fails, HTTP status is still typically 200 and this object is returned with result.isError: true. Use that flag rather than HTTP status alone.
2.0 Request identifier echoed back
MCP tool result. Inspect isError for tool-level success or failure.
isError: true — the tool failed; content usually has one text item with a human-readable explanation. Typical causes:
- Invalid arguments (e.g. empty
prompt, invalid UUID forchat_id/insight_idwhere required). - Downstream errors (e.g. resource not found, timeout, cancelled request, or generic internal failure).
_meta.isInternalError — when present and true, the server classified the failure as internal or timeout-style. Omitted or false for most validation and not-found style errors.
This is not the same as a JSON-RPC protocol error: invalid method or malformed requests use the standard JSON-RPC error field instead of a tool result.
Show child attributes
Show child attributes
Was this page helpful?

