Gemini API in India — ₹-Native Access

Google's Gemini models are reachable through unoblox's OpenAI-compatible endpoint in India — one key, one GST invoice, and pricing quoted straight in rupees rather than routed through a foreign billing account.

Gemini in India without a separate Google Cloud account

Normally, using Gemini means setting up Google Cloud billing, choosing a region, and paying through a foreign-billed account. On unoblox you skip all of that: point your existing OpenAI SDK at https://api.unoblox.ai/v1, use your ub-gw-… key, and Gemini sits alongside every other model on the same invoice.

That matters most for teams that already call Gemini directly and want a second, India-billed path — for example, to compare Gemini's vision handling against GPT-4o or Claude Sonnet on the same document set without provisioning a second cloud account just to run the test.

Why run Gemini through unoblox

• One endpoint for every model — DeepSeek, Qwen, Claude, GPT, and Gemini all answer to the same base URL.
• Rupee-only billing — a single monthly GST invoice, input-tax credit claimable.
• Drop-in SDK compatibility — swap the model string from GPT or Claude to Gemini without touching the rest of your integration.
• Multimodal by default — Gemini's strength on images, PDFs, and video comes through unchanged.

Pricing

Gemini pricing on unoblox is live on /models rather than fixed here, since rates move as new Gemini tiers are added — check the model page for the current ₹ input/output rate on each variant before you commit a workload to it.

How to call it

from openai import OpenAI

client = OpenAI(api_key="ub-gw-...", base_url="https://api.unoblox.ai/v1")
response = client.chat.completions.create(
    model="google/gemini-pro",  # see /models for the exact live model id
    messages=[{"role": "user", "content": "Summarize this document..."}],
)

Frequently asked questions

Can I use Gemini's vision features on unoblox? Yes — pass images or PDF pages in the messages array the same way you would with Google's own API; multimodal handling comes through the gateway unchanged.

Do I need a Google Cloud account? No. Sign up on unoblox with your email, generate a key, and you're calling Gemini within minutes.

Is Gemini priced the same on unoblox as directly with Google? unoblox quotes its own rupee rate per model on /models — check there for the current number rather than assuming parity with any other source.

What about rate limits? Every key has a requests-per-minute tier you can raise from the dashboard; new accounts start with free credits to test before committing spend.

Can I use Gemini with LangChain or LlamaIndex? Yes — both frameworks talk to any OpenAI-compatible base URL, so pointing them at unoblox works without custom integration code.

Which Gemini variant should I start with? Start with whichever tier /models marks as the fast, cost-efficient option for simple tasks, and move to the higher-end tier only when a task genuinely needs deeper reasoning or heavier multimodal input.

Get started in rupees → https://unoblox.ai/sign-in