API429 in Make: HTTP v4 keychain credentials and a manual request

Create an unscheduled scenario with HTTP v4 Make a request

Create a dedicated test scenario in Make and keep Scheduling switched Off. Add the HTTP Version 4 app and select the Make a request action. Older HTTP module versions expose different credential controls, so confirm that you are using HTTP v4 in your workspace before entering settings. A single HTTP module triggered manually with Run once is sufficient to verify authentication, routing and response parsing.

Do not attach a schedule, webhook, iterator or automated error-handler retry loop during your initial check. Automatic retries or multi-bundle iterators can multiply paid model calls before you have verified the response structure. Keep the test prompt short and nonconfidential because Make records module inputs and outputs in scenario execution history according to your organization settings.

Distinguish the request JSON example from a Make blueprint

This guide provides an explicit field list and the literal HTTP request body JSON rather than an importable Make scenario blueprint. A Make blueprint file contains application-specific module identifiers, visual coordinates, router metadata and connection references that depend on your workspace state. Do not attempt to import the Chat Completions JSON payload below through Make's Import Blueprint menu.

Keeping the HTTP request body separate lets you inspect every field sent over the wire: model, messages, max_tokens and stream. If you prefer an importable two-node file for another automation tool, see the inactive manual workflow on the n8n integration page (/en/integrations/n8n); in Make, enter the module fields manually and paste the JSON payload into the HTTP v4 request body.

Store Bearer authentication in Credentials and keychain

In the HTTP v4 module, set Authentication type to API key, open Credentials and create a dedicated keychain entry. In the Key field, enter Bearer YOUR_API429_KEY, replacing only the placeholder with your real API429 client key inside Make's protected credential dialog. Set API key placement to Header and set API key parameter name to Authorization. Because the Bearer prefix is already part of the Key value, do not add a second Bearer prefix.

Do not place your API key in the ordinary Headers list of the HTTP module, and never embed it in the URL, query string, JSON request body or scenario notes. Storing the secret in Credentials/keychain keeps it in Make's encrypted credential store rather than exposing plain-text tokens in module configuration screenshots or exported scenario definitions. Verify GET /v1/models with your key separately, as saving a keychain entry does not prove model access.

Configure the Chat Completions body and inspect mapper output

Set Method to POST and URL to https://gateway.api429.com/v1/chat/completions. Choose application/json as the body content type and paste the JSON body with model set to gemini-3.8-flash, a single user message, max_tokens set to 256 and stream set to false. Even though gemini-3.8-flash belongs to the Google family, this route uses the OpenAI-compatible Chat Completions envelope rather than Google's native GenerateContent payload.

Click Run once to send one deliberate test request—which can incur paid API usage—and inspect the returned HTTP status and parsed output bundle. In standard JSON notation, the assistant reply sits at choices[0].message.content. Because Make's visual mapper displays collections and array indices with its own visual tokens (often starting at 1 in the picker UI), select the actual message.content field shown in your execution output rather than typing raw bracket syntax blindly.

Diagnose HTTP errors and separate Make operations from token spend

If the module returns 401, check the selected keychain credential, confirm Header placement under the name Authorization and ensure there is exactly one Bearer prefix followed by a space. For a 404 error, verify the full URL https://gateway.api429.com/v1/chat/completions without a duplicate /v1. For a model error, compare the literal ID gemini-3.8-flash against the authenticated GET /v1/models list for your key. If you hit 429, stop and inspect rate limits instead of enabling retries.

Keep Make's own operation counting separate from API429 token billing. Each module execution consumes Make operations under your Make plan, while each model request consumes Google family input and output tokens in API429. The max_tokens: 256 field caps only the completion length of that single HTTP call, not total scenario spend. Leave Scheduling set to Off until you have measured real token usage and bounded how many bundles a production trigger can pass.

HTTP v4 module fields for manual entry

Module: HTTP v4 > Make a request
Method: POST
URL: https://gateway.api429.com/v1/chat/completions
Authentication type: API key
Credentials > Key: Bearer YOUR_API429_KEY
API key placement: Header
API key parameter name: Authorization
Body content type: application/json
Scheduling: Off
Run: Run once

Request body JSON, not a Make scenario blueprint

{
  "model": "gemini-3.8-flash",
  "messages": [
    {
      "role": "user",
      "content": "Answer in one sentence: what is an API?"
    }
  ],
  "max_tokens": 256,
  "stream": false
}

Payment and a $10 test

• Russian bank card. Top up your balance in the account.
• Invoice for companies. Bank transfer with closing documents.
• USDT and TON. Top up in crypto.

$10 test credit. Subscribe to our Telegram channel through the bot and get $10 on your balance for the first requests. Get $10 in the bot