Create an account
Sign up with email in about 30 seconds. No card needed.
Pay per token in ₹
Know the cost before you spend.
Pay exactly for what you use, in rupees, not dollars. No credit card required. Top up via UPI, net banking, or a domestic debit card.
- No subscriptions
- No minimums
- Top up from ₹100
Live estimateClaude Sonnet 4.6
₹142.68
for 200K input and 50K output tokens
Example: Claude Sonnet 4.6. Full live pricing is on the models page.
HOW A REQUEST WORKS
One key. Many models.
One balance.
Each request goes to the right model through one API key, and the answer comes back in one response.
EXAMPLE REQUEST
Summarize Tickets
POST /v1/chat/completions
- Send the ticket
- Pick a model
- Read the summary
A two line summary with an urgency tag.
zenntraa
MODELS USED
NNova LiteAmazon
LLlama 4 ScoutMeta
2 models · 2 calls · paid from one balance
0/2 calls back·waiting on modelsexample charge ₹0.01, from one balance
One key gives you access to every model.
Top up in rupees and pay only for what you use.
Switch in two lines
Every OpenAI-compatible SDK works as is. Change the base_url and api_key, and keep the rest of your code.
from openai import OpenAI client = OpenAI( base_url="https://api.zenntraa.com/v1", api_key="sk-your-key", # that's the whole migration) response = client.chat.completions.create( model="claude-sonnet-4-6", messages=[{"role": "user", "content": "Hello!"}], stream=True,) for chunk in response: print(chunk.choices[0].delta.content or "", end="")How it works
Running in three steps
Top up in rupees
Add ₹100 or more via UPI, net banking, or a debit card.
Call any model
Change base_url and api_key in your code. Everything else stays the same.
Prompts not stored
We log model, tokens, cost, and time. Never your content.
Balance never expires
Your rupees stay put until you use them.
Refund unused balance
Email us and we refund whatever is left.
No international card
Pay in ₹ through UPI or a domestic card.
Built for
Who uses Zenntraa
Students
Build AI projects without an international credit card. Start with a small top-up.
Startups
One key and one ₹ wallet instead of managing several provider accounts.
Developers
Call Claude, GPT, Gemini, and DeepSeek from the same OpenAI-style code.
Teams
Set per key rate limits so each project or script has its own ceiling.
Everything included
Drop-in compatible
Point your existing SDK at our base_url and go.
One key, many models
44+ models across Anthropic, DeepSeek, Mistral, Meta, Amazon, OpenAI, Google, NVIDIA, MiniMax, Writer, Moonshot, Alibaba, and Z.AI behind one key.
INR wallet
Top up via UPI, net banking, or a domestic debit card.
Spend controls
Per key rate limits so a runaway script never drains your wallet.
Streaming
Responses stream token by token, same as calling the provider directly.
Live rupee pricing
Live USD to INR conversion. You see the exact rupee cost before you spend.
Frequently asked questions
Quick answers before you top up.
01Do you store my prompts?
No. We only log model, token count, cost, and timestamp. Your prompt content and responses never touch our storage.
02What happens if my wallet hits zero mid-request?
The request goes through if it was already in flight. Your balance can go slightly negative. Just top up and it adjusts.
03What if a request fails on the model side but I still got charged?
Email us and we refund it. No questions asked.
04Does my balance expire?
No. Your rupees sit there until you use them.
05Can I get a refund on unused balance?
Yes. Email us and we'll refund whatever is left.
06Is this legal to use in India?
Yes. You're paying for a software service in INR. No international transaction, no RBI issue.
07How much latency does the gateway add?
Typically under 50ms overhead. Most of your latency is the model itself.
Get started
Your first API call is minutes away
Create an account, top up with UPI, and call any model using the code you already have.
- UPI
- Net banking
- Debit card
- ₹100 minimum