GPT-5 Mini Pricing in India (₹ per Token)

GPT-5 mini is OpenAI's smaller, faster sibling to GPT-5, built for high-volume tasks where full GPT-5 reasoning would be overkill. On unoblox it is billed the same way every other model is: in rupees, per token, on one monthly GST invoice — no international card and no overseas invoice to reconcile.

GPT-5 mini pricing on unoblox

Model	Input (₹ / 1M tokens)	Output (₹ / 1M tokens)
GPT-5 mini	₹25.2	₹201.6
GPT-5 (for comparison)	₹126	₹1,008

GPT-5 mini costs roughly a fifth of full GPT-5 on both input and output. For workloads where mini's quality is good enough — classification, extraction, short-form chat, drafting — that difference compounds fast at volume.

A worked example

A request with 2,000 input tokens and 800 output tokens on GPT-5 mini costs: (2,000/1,000,000 × ₹25.2) + (800/1,000,000 × ₹201.6) = ₹0.050 + ₹0.161 ≈ ₹0.21. At 50,000 such requests a month, that is roughly ₹10,500 — a number you can plan around directly in rupees rather than back-calculating from a rate card priced in something else.

Why teams reach for a "mini" model

• High request volume: chatbots, autocomplete, and classification pipelines that fire thousands of times a day benefit most from a lower per-token rate.
• Latency-sensitive paths: smaller models generally respond faster, which matters for anything user-facing.
• A cost-aware default: many teams route everything to a mini-class model first and escalate only the requests that need more reasoning to a larger model like GPT-5.

Calling GPT-5 mini from your existing stack

No new SDK required — point your current OpenAI client at unoblox and change two values.

base_url: https://api.unoblox.ai/v1
api_key: ub-gw-xxxxxxxxxxxxxxxxxxxx
model: openai/gpt-5-mini

Where GPT-5 mini sits in the GPT-5 family

unoblox's GPT-5 line spans openai/gpt-5, openai/gpt-5-mini, openai/gpt-5-nano, the gpt-5.1, gpt-5.2, gpt-5.4 and gpt-5.5 point releases, and the gpt-5.6 family (luna, sol and terra variants). Only GPT-5 and GPT-5 mini have a published indicative rate today; for the rest, check the live rate on the model's own page, for example /models/openai/gpt-5-nano, or browse the full catalog at /models.

Frequently asked questions

Is GPT-5 mini good enough to replace GPT-5 for most tasks? For well-scoped tasks like classification, tagging, and short drafts, many teams find mini-class models sufficient. For multi-step reasoning or long-context work, full GPT-5 is usually worth the higher rate. This is a per-workload decision, not a universal answer.

Does unoblox charge anything on top of the per-token rate? The per-token ₹ rate shown on the model page is what you pay for usage; billing rolls up into one monthly GST invoice rather than multiple line items per provider.

Can I switch between GPT-5 mini and GPT-5 without changing my integration? Yes — both are called through the same OpenAI-compatible endpoint and key. Changing the model field in your request is the only change needed.

Is GPT-5 mini pricing on unoblox different from OpenAI's own pricing? unoblox publishes its own indicative ₹ rate per model, shown live on the model's page, billed in rupees on a GST invoice — that is the main practical difference for an India-based account.

Does GPT-5 mini run on infrastructure located in India? No — GPT-5 mini is an OpenAI model and runs on OpenAI's infrastructure. unoblox's India benefit here is rupee billing, GST invoicing, and one endpoint — not data residency.

Where do I find the live rate if it changes? Every model's current rate is shown on its own catalog page, for example /models/openai/gpt-5-mini — treat the rates on this page as indicative and the model page as the source of truth.

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