Gemini 3.8 Flash and Gemini 3.1 Pro are now half price on Unoblox

Gemini 3.8 Flash and Gemini 3.1 Pro Preview are now 50% off on Unoblox. Both are available through the same OpenAI-compatible API you already use, with a million-token context window, tool calling and structured outputs, billed in rupees. This is launch pricing and runs while our promotional Google Cloud credits last. See Gemini 3.8 Flash

The new prices

Rates are per 1 million tokens, observed in the Unoblox catalog on 5 October 2026.

Model	Input (full → now)	Output (full → now)	Cached input (now)
Gemini 3.8 Flash	₹75.76 → ₹37.92	₹378.86 → ₹189.61	₹3.79
Gemini 3.1 Pro Preview	₹202.06 → ₹101.12	₹1,212.34 → ₹606.77	₹10.11

Gemini 3.1 Pro Preview has a higher tier when a single request uses 2,00,000 or more prompt tokens: input ₹202.26 and output ₹910.16 per million at the discounted rate, applied to that whole request. Output billing includes reasoning tokens for both models. Applicable platform fees and GST are additional; the live model pages are authoritative.

Which Gemini should you use?

Gemini 3.8 Flash is the volume option. At ₹37.92 per million input tokens, it suits high-throughput work: classification, extraction into JSON, summarising support tickets, routing and first-pass agent steps. Streaming is supported.

Gemini 3.1 Pro Preview is the heavier reasoning option for harder tasks: multi-step analysis, long-document questions and complex tool-using agents. Use it where Flash's answers are not good enough for your task.

A practical pattern is to start every workload on Flash, measure quality on a sample of your own data, and promote only the cases that need it to Pro. These are suggested uses based on the published model profiles, not Unoblox benchmark results.

What both models bring

Capability	Gemini 3.8 Flash	Gemini 3.1 Pro Preview
Context window	10,48,576 tokens	10,48,576 tokens
Input	Text, image	Text, image
Structured outputs (JSON schema)	Yes	Yes
Tool calling	Yes	Yes
Streaming	Yes	Not documented

Capabilities as listed on the Unoblox model pages on 5 October 2026. Both routes are served by Ogma through Google Vertex AI. Zero-data-retention routing is available, and your data is never used for training under platform defaults.

Start building

Point any OpenAI SDK at https://api.unoblox.ai/v1 and choose a model ID:

import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://api.unoblox.ai/v1",
  apiKey: process.env.UNOBLOX_API_KEY,
});

const res = await client.chat.completions.create({
  model: "google/gemini-3.8-flash", // or "google/gemini-3.1-pro-preview"
  messages: [{ role: "user", content: "Summarise this ticket in one line." }],
});

New to Unoblox? Your first wallet top-up is matched up to ₹250, so the half-price rates stretch further. Explore Gemini on Unoblox