
How to Pay for Gemini API in India (UPI, No International Card)
Step-by-step guide for Indian developers to access Google's Gemini models and pay in rupees via UPI or net banking — no international credit card, no USD billing.
Author
AICredits Team
Published
7 Jul 2026
Reading time
6 min read
The problem: Google bills Gemini API usage in USD
Google AI Studio hands out free-tier Gemini keys, but the moment you need production rate limits you are on paid billing — which runs through a Google Cloud billing account, charges in USD, and expects a card that works for international transactions. For many Indian developers that means forex conversion fees on every invoice, unpredictable rupee amounts month to month, and the familiar card-declined error when a domestic debit card meets an international charge.
AICredits removes that entire step for the Gemini API in India. You top up a prepaid wallet in INR via UPI, net banking, or a domestic debit card, and Gemini usage is deducted from that wallet in rupees. No Google Cloud billing account, no USD statement, no international card.
Step 1: Create your AICredits account
Sign up at aicredits.in. No credit card is required at signup, and the dashboard is available immediately.
Navigate to API Keys and click Create New Key. Name it something like gemini-production and optionally set a budget cap so a runaway script can never spend more than you allow.
Step 2: Add credits in INR
Go to Billing → Add Credits. The minimum topup is ₹10. Pay via Razorpay using UPI (Google Pay, PhonePe, Paytm), net banking, or a domestic debit card.
Credits land in your wallet instantly after payment confirmation. Every Gemini call you make afterwards deducts its exact cost in rupees from this balance — you can watch it in the usage dashboard per request.
Step 3: Point your code at AICredits
AICredits exposes an OpenAI-compatible endpoint, so you use the OpenAI SDK with a Gemini model ID. Two lines change:
from openai import OpenAI
client = OpenAI(
base_url="https://api.aicredits.in/v1",
api_key="sk-your-aicredits-key",
)
response = client.chat.completions.create(
model="google/gemini-2.5-flash",
messages=[{"role": "user", "content": "Summarise this support ticket in two lines."}],
)
print(response.choices[0].message.content)Streaming, system prompts, JSON output, and tool calling work as documented. You can also write the model as plain gemini-2.5-flash — the gemini- prefix routes to Google automatically.
What Gemini costs in rupees
Standard-tier prices through AICredits, per million tokens (includes forex conversion and platform markup — the number you see is the number your wallet pays):
| Model | Input (₹/1M) | Output (₹/1M) | Best for |
|-------|--------------|---------------|----------|
| google/gemini-2.5-pro | ~₹120 | ~₹963 | Complex reasoning, long documents, code |
| google/gemini-2.5-flash | ~₹29 | ~₹241 | Production default — fast and capable |
| google/gemini-2.5-flash-lite | ~₹9.6 | ~₹38.5 | Classification, extraction, high volume |
| google/gemini-2.0-flash | ~₹9.6 | ~₹38.5 | Cost-efficient general tasks |
A typical chatbot exchange — 500 input tokens, 300 output tokens on Gemini 2.5 Flash — costs about ₹0.09. A ₹500 topup covers roughly 5,500 such exchanges. Prices track the live forex rate, so check aicredits.in/calculator for current numbers against your own workload.
Gemini 2.5 models also support cached input at a fraction of the standard input price, which matters if you send the same long system prompt or document on every request.
Why not just use the Google AI Studio free tier?
The free tier is genuinely useful for prototyping, and nothing stops you from starting there. The gap appears in production:
- Free-tier rate limits are low and shared per project — fine for a demo, not for real traffic.
- Paid billing requires the Google Cloud + international card setup described above.
- There is no per-key budget control, so one misbehaving loop bills your card directly.
With a prepaid INR wallet, spend is capped by design: the wallet balance and per-key budgets are hard limits. When credits run out, requests stop — your card is never on file to overdraw.
Frequently Asked Questions
Can I pay for the Gemini API with UPI?
Not directly with Google. Through AICredits, yes — topups go through Razorpay, which accepts UPI (Google Pay, PhonePe, Paytm), net banking, and domestic debit cards. Your Gemini usage then deducts from the INR balance.
Do I need a Google Cloud account?
No. AICredits holds the provider relationship with Google. You only need an AICredits account and an API key.
Which Gemini models are available?
All major current models: Gemini 2.5 Pro, 2.5 Flash, 2.5 Flash-Lite, and the 2.0 Flash family. The full list with live rupee pricing is at aicredits.in/models.
Does streaming and tool calling work?
Yes. The OpenAI-compatible format supports streaming, tool/function calling, JSON mode, and vision inputs for Gemini models, translated to Google's API internally.
Get started
- Create a free account and top up ₹10 via UPI to make your first Gemini call.
- Follow the quickstart guide — two lines of config and you are live.
- Estimate your workload on the cost calculator, or see the Gemini API India overview.
- Also using OpenAI models? Read How to Pay for OpenAI API in India Without an International Credit Card.
Related Articles
Continue in Docs
Need implementation commands and endpoint details? Go to quickstart or API reference.