Documentation

The complete Vantenza AI guide

Everything you need to use your Vantenza AI API key: from buying quota and connecting it to apps, to example prompts for your job.

What is Vantenza AI?

Vantenza AI is an AI (artificial intelligence) service that can help you write, summarize, translate, analyze data, and even write code, in Indonesian as well as other languages.

Pay for what you need

No monthly subscription. You buy quota once, then use it until it runs out.

No credit card

Pay through iPaymu on commerce.vantenza.com (QRIS, e-wallets, virtual accounts, and more).

Text, images, video & files

Send questions, photos, videos, or PDF documents, and use it in a chat app, Google Sheets, or your own program with one API key.

Start here

Follow these three steps in order. Each step has a full guide below.

  1. 1

    Get an API key

    Sign up or sign in on commerce.vantenza.com, then buy quota. Your API key is saved right in your account.

    Buy quota
  2. 2

    Use the AI

    Enter your API key into a chat app, Google Sheets, or your program, then start asking questions.

    See how to use it
  3. 3

    Watch & top up

    See your remaining quota and usage in your account dashboard, then top up the same key when it runs low.

    Open the dashboard

1. Get an API key

Your API key is created in your account on commerce.vantenza.com as soon as your first quota purchase is paid.

  1. 1Create an account on commerce.vantenza.com with your email and a password, or sign in if you already have one.
  2. 2Click "Buy quota", choose an amount and "New key", then pay on the iPaymu page with QRIS, an e-wallet, or a virtual account.
  3. 3Once the payment is confirmed, open https://commerce.vantenza.com/account/ai. Your new API key is there, with its quota.
  4. 4Click "Show the key", enter your account password again, then copy the key. The Base URL and model name are in the "Access details" section below.
  5. 5Keep your API key somewhere safe. You can show it again anytime from the dashboard.

2. The easiest way to use it

Never used an API before? No problem. The easiest way is to use an AI chat app that supports the "OpenAI API", then fill in three details: the Base URL, the API key from your account dashboard, and the model name.

  1. 1Install an AI chat app that has an "OpenAI API Compatible" / "Custom API" setting (available for Windows, Mac, Android, and iPhone).
  2. 2Open the Settings menu in that app, then choose the "OpenAI API Compatible" or "Custom" provider.
  3. 3Set API Host / Base URL to: https://api.vantenza.com
  4. 4Set API Key to the API key from your account dashboard.
  5. 5Set Model to: Vantenza-V2.6 (or another model from the "Model options" section)
  6. 6Save, then type your first question, for example "Hi, what can you help me with?"
text
Provider : OpenAI API Compatible
Base URL : https://api.vantenza.com
API Key  : vt-YOUR-API-KEY
Model    : Vantenza-V2.6

Work in spreadsheets more often? Use the =VANTENZA() function in Google Sheets. The step-by-step guide is in the "All ways to connect" section → Google Sheets.

See the Google Sheets guide

3. Watch your quota & top up

Your quota goes down every time you use the AI. Watch what is left and add quota from your account dashboard.

  1. 1Sign in to your account, then open https://commerce.vantenza.com/account/ai.
  2. 2See each API key's total quota, quota used, and remaining quota (in rupiah).
  3. 3Click "Show usage" to see the cost, requests, and tokens per day and per model.
  4. 4Running low? Click "Top up" on that key. The quota grows and the API key stays the same, so your apps need no change.
  5. 5Need a separate key, for example for another app? Buy a new key. "Show the key", "Replace the key", and "Revoke the key" ask for your account password.

Access details

Use the following values in any app or code that supports the OpenAI format.

Base URLhttps://api.vantenza.com
Some apps ask for a /v1 suffix; both work.
Chat endpointhttps://api.vantenza.com/chat/completions
POST method, OpenAI Chat Completions format.
ModelVantenza-V2.6
The default model. Other models are in the "Model options" section.
API keyvt-YOUR-API-KEY
From your account dashboard. Send it in the header Authorization: Bearer <API key>.
Dashboardhttps://commerce.vantenza.com/account/ai
Remaining quota, usage per day and model, and top-ups.
bash
curl https://api.vantenza.com/chat/completions \
  -H "Authorization: Bearer vt-YOUR-API-KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "Vantenza-V2.6",
    "messages": [
      {"role": "system", "content": "You are an assistant who answers briefly in English."},
      {"role": "user", "content": "Explain what inflation is in 2 sentences."}
    ]
  }'

Model options

One API key can use every model below. Just change the model name in your app. This list always shows the models that are currently active.

Chat

Vantenza-V2.6-Lite

Paling hemat dan tercepat. Cocok untuk chat, livechat, Google Sheets, dan membaca gambar.

teksgambarvideofile
Max. answer 32,768 tokensMax. request size 20 MB
Vantenza-V2.6-Pro

Paling pintar. Cocok untuk coding kompleks, agent, dan analisis berat.

teksgambarfile
Max. answer 32,768 tokensMax. request size 20 MB
Vantenza-V2.6-Reasoning

Berpikir bertahap. Cocok untuk matematika, logika, dan coding yang butuh penalaran, lebih hemat dari Pro.

teks
Max. answer 32,768 tokensMax. request size 4 MB

Specialized models

Vantenza-Embed
Embedding (text search)

Mengubah teks jadi vektor untuk pencarian dan RAG.

teks
Max. request size 4 MB
Vantenza-Image
Image generation

Membuat gambar dari teks.

teks
Max. request size 4 MB
Vantenza-Voice
Text to speech

Mengubah teks jadi suara.

teks
Max. request size 1 MB
Vantenza-Transcribe
Speech to text

Mengubah rekaman suara jadi teks.

audio
Max. request size 20 MB

Change the model name in your app to use a different model. Smarter models use up quota faster.

Non-chat models are used through their own endpoints, with the same quota. See examples

All ways to connect

Pick the option that suits you best. Every option uses the same Base URL, model, and API key. If you are not a programmer, start with "Chat apps & plugins" or "Google Sheets".

The fastest way to test your API key from a terminal.

cURL (terminal)

Send a single request straight from Terminal (macOS/Linux) or PowerShell/Git Bash (Windows). Handy for making sure your API key is active.

Steps

  1. 1Copy the command below.
  2. 2Replace vt-YOUR-API-KEY with the API key from your account dashboard.
  3. 3Run it in the terminal. The AI's answer is in choices[0].message.content.
bash
curl https://api.vantenza.com/chat/completions \
  -H "Authorization: Bearer vt-YOUR-API-KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "Vantenza-V2.6",
    "messages": [
      {"role": "system", "content": "You are an assistant who answers briefly in English."},
      {"role": "user", "content": "Explain what inflation is in 2 sentences."}
    ]
  }'

More API features

Besides chat, the same API key works for embeddings, image generation, text to speech, speech to text, and balance checks. All usage is deducted from the same rupiah quota. The examples below use models that are active right now.

Endpoint list

For programmers whose library or tool uses a different format.

Chat: 3 formats

Every chat model can be called in three formats: OpenAI Chat Completions (examples in "All ways to connect"), OpenAI Responses, and Anthropic Messages. Pick the one your library supports; results and quota are the same.

Steps

  1. 1Chat Completions: https://api.vantenza.com/v1/chat/completions. The most common format, used by almost every app.
  2. 2Responses: https://api.vantenza.com/v1/responses. The answer is in output[].content[].text.
  3. 3Messages (Anthropic format): https://api.vantenza.com/v1/messages. Send the API key in the x-api-key header or as Authorization: Bearer. The answer is in content[0].text.
  4. 4Replace vt-YOUR-API-KEY with your API key, then run it in a terminal.
bash
curl https://api.vantenza.com/v1/responses \
  -H "Authorization: Bearer vt-YOUR-API-KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "Vantenza-V2.6",
    "input": "Explain inflation in 2 sentences."
  }'

Guides by job

Example uses and ready-to-use prompts for different jobs. Click "Copy", then replace the parts in [square brackets].

Programmer

Add AI capabilities to your apps, code faster, and automate routine work.

Can be used for

  • Building AI features into apps: chatbots, automatic summaries, ticket classification, data extraction.
  • Explaining, reviewing, and cleaning up code.
  • Writing unit tests, documentation, and commit messages.
  • Writing SQL queries and regexes from plain-language descriptions.
  • Turning open-ended text into structured JSON.

Recommended options

PythonJavaScript / Node.jsPHP / LaravelcURL (terminal)Chat apps & plugins

Tips

  • Use a "system" role message to set the role and answer format consistently.
  • Ask for a structured format ("reply only with JSON") so the result is easy for programs to process.
  • Answer length per request is limited by the model (see "Model options"). For long results, split the work into several requests.
  • Store your API key in an environment variable or secret manager, not in your repository.

Example prompts

Code review

You are a senior software engineer. Review the following code: look for bugs, security issues, and parts that could be simplified. Present them as a numbered list along with the fixes. ``` [paste code] ```

Unit tests

Write complete unit tests (normal, edge, and error cases) for the following function using [pytest / Jest / PHPUnit]. Include a short explanation of each test. ``` [paste function] ```

Text to JSON

Extract data from the following text into JSON with the fields: name, email, phone, city. Use null if a value is missing. Reply ONLY with valid JSON. Text: [paste text]

SQL query

Tables: orders(id, customer_id, total, created_at), customers(id, name, city). Write a PostgreSQL query that shows the 10 cities with the highest total sales last month, then briefly explain how it works.

Commit message

Write a commit message in the Conventional Commits format (max. 72 characters for the title) from the following changes: [paste git diff]

Quota & tokens

Understand how quota is counted so you can use it more efficiently.

Quota in rupiah

Your API key quota equals the amount you paid. Each request reduces the quota based on the number of tokens processed.

What is a token?

Tokens are pieces of words. Roughly 1 token ≈ 4 characters of English; Indonesian usually takes slightly more tokens per word.

Input & output

Input is all the text you send (including the conversation history). Output is the AI's answer. Output tokens cost more, so ask for concise answers when you can.

Limits per request

Each model has its own limits on answer length and request size (see "Model options"). Images, video, and files also count as input tokens.

Tips for saving quota

  • Ask for short, specific answers, for example "3 points max" or "answer in one word".
  • Send only the relevant part of a document, not the whole thing.
  • Limit the conversation history: send only the last few messages.
  • Save (cache) answers to repeated questions so they are not sent again and again.

Prompt writing tips

The quality of the answer depends a lot on how you ask. Use the following pattern.

Give it a role

Start with the AI's role, for example "You are an experienced accountant".

Explain the context

Mention the goal, the audience, and any relevant background information.

Specify the format

Ask for the shape of the answer: bullet points, a table, JSON, or a certain number of words.

Give examples

One or two examples of the result you want make answers much more consistent.

Break up big tasks

Ask for an outline first, then develop it section by section.

Iterate

Refine with follow-up messages: "shorter", "more formal", "add examples".

Prompt template

text
Role    : You are [a role, for example an experienced financial analyst].
Context : [background and goal].
Task    : [what needs to be done].
Format  : [bullet points / table / JSON / max N words].
Data    :
[paste data or text here]
✗ Unclear

Write a caption.

✓ Clear

You are a social media copywriter. Write 3 Instagram captions (max. 30 words, casual style) for a 20% off promo on palm-sugar iced coffee, aimed at college students. End with a call to buy.

If you get an error

If a request fails, the API returns an HTTP status and an error message in the following format:

json
{
  "error": {
    "message": "Your quota has been used up. Please purchase more quota.",
    "type": "insufficient_quota",
    "code": "quota_exhausted"
  }
}
StatusMessageCauseSolution
401Invalid API key.The API key is wrong, incomplete, was not sent, or was replaced or revoked.Copy the API key again from your account dashboard. Make sure the header is in the format "Authorization: Bearer <API key>".
429Your quota has been used up. Please purchase more quota.The API key's quota has run out.Check your remaining quota at https://commerce.vantenza.com/account/ai, then top up the same key. The API key does not change.
429The service is busy. Please retry in a moment.Too many requests in a short time.Wait a few seconds, then try again. In programs, add a delay and retry automatically (retry with backoff).
413Request is too large for this model (max …).The request content (text, conversation history, images, video, files) exceeds the model's limit.Shrink or compress the file, send the document in parts, trim the conversation history, or use a model with a larger limit.
400Invalid model name passed in model=…The model name is wrong, or that model is currently inactive.Use a name from the "Model options" section, for example "Vantenza-V2.6". Model names are case-sensitive.
404Not found.The endpoint address is wrong.Use https://api.vantenza.com/chat/completions (or the Base URL https://api.vantenza.com in libraries).
500 / 503The service is temporarily unavailable. Please retry in a moment.A temporary problem on the server.Try again in a little while. If it continues, contact cs@vantenza.com.

Security

An API key is just like a password: anyone who has it can use your quota.

  • Do not share your API key with other people or in group chats.
  • Do not put your API key in code that runs in the browser, in mobile apps, or in public repositories.
  • Store your API key in an environment variable, a secret manager, or Script Properties (Google Sheets).
  • If your API key leaks, use "Replace the key" in the dashboard right away: the old key stops working at once and the quota stays. Need help? Contact cs@vantenza.com.
  • Do not send sensitive data (passwords, card numbers, other people's personal data) to the AI.
  • AI answers can be wrong. Double-check facts, numbers, and code before using them.

Glossary

Terms that come up often, explained in everyday language.

Top-up
Adding quota to an API key you already have. The key stays the same, so your apps need no change.
API key
A personal key (like a password) that lets the AI recognize you and deduct from your quota. Do not share it.
Base URL
The Vantenza AI server address you enter into apps, which is https://api.vantenza.com.
Model
The name of the AI "brain" being used, for example Vantenza-V2.6. The full list of options is in the "Model options" section.
Prompt
The instruction or question you type for the AI.
Token
The unit of word pieces the AI counts. The longer the text, the more tokens are used.
Quota
Your usage balance (in rupiah), which goes down according to the number of tokens used.
Endpoint
The full address for sending questions from a program: https://api.vantenza.com/chat/completions.

FAQ

Frequently asked questions.

What is an API key, and what is it for?

An API key is a secret code that identifies you when an app uses the AI. Paste it into a chat app, Google Sheets, or your code together with the Base URL and model name, and every use is counted against your quota. Keep it safe and do not share it with anyone.

How do I get an API key?

Sign up or sign in on commerce.vantenza.com, then buy Vantenza AI quota. Once the payment is confirmed, the API key shows in your account dashboard (https://commerce.vantenza.com/account/ai) with its quota.

I lost my API key. How can I see it again?

Sign in to your account, open https://commerce.vantenza.com/account/ai, then click "Show the key" and enter your account password. If the key may have leaked, use "Replace the key": the old key stops working and the quota stays.

How do I check my remaining quota?

Open https://commerce.vantenza.com/account/ai. Your remaining quota is shown in rupiah, and "Show usage" lists the cost, requests, and tokens per day and per model.

What is an AI token, and how much can I use with my quota?

Quota is counted by the number of tokens (pieces of words) processed: the text you send (input) and the AI's answer (output). Answers cost more than input. The tokens you have used are shown per day and per model in your account dashboard.

What happens when my quota runs out?

Your next requests will be declined with the message "Your quota has been used up". Top up the same key to carry on; the API key does not change.

Does the quota expire?

Your quota does not go down if you do not use it. Contact cs@vantenza.com if you have questions about the purchase terms.

Can one API key be used in several apps?

Yes. All usage comes out of the same quota. Do not share your API key with anyone else.

Does the AI remember earlier conversations?

No. Each request stands on its own. To continue a conversation, send the earlier message history again in the messages array (see the "Conversation" example in the Python section). Chat apps usually do this automatically.

Why is the AI's answer cut off?

Each model has a limit on answer length (see "Model options"). If the answer is cut off (finish_reason = "length"), ask the AI to keep going with the message "continue", or split the task into several parts.

Can I send images, video, or files?

Yes. Send photos, videos, or PDF documents along with your question (examples are in the "Images, video & files" tab). The input types and maximum size for each model are listed in the "Model options" section. Generating images and speech is available through separate API endpoints (see the "More API features" section of the guide); generating video is not supported yet.

Which model should I use?

Start with the default model. Use a "Lite" model for light chat and high volumes to save quota, and a smarter model for coding or complex analysis. All models use the same quota; smarter models use it up faster.

Is my data safe?

Avoid sending sensitive data such as passwords, card numbers, full bank account numbers, or other people's personal data. Anonymize data before sending it.

Need help? Contactcs@vantenza.com