Qwen3 1.7B free API in India

Qwen3 1.7B is the one model on unoblox priced at ₹0 — a genuinely free tier, callable from India through the same OpenAI-compatible endpoint and ub-gw- key as every paid model in the catalogue, with no card and no invoice needed to start.

Why a free model exists in a ₹-billed catalogue

Most of unoblox exists to make paid inference billable in rupees with a proper GST trail. Qwen3 1.7B is the exception on purpose: a small, fast model that costs nothing to call, so a new account can build and test an integration — get the request shape right, wire up streaming, validate a prompt — before spending a single rupee on a larger model. It's also genuinely useful on its own for lightweight tasks that don't need a large model's reasoning depth.

₹ pricing for Qwen3 1.7B

Model	Input (₹ / 1M tokens)	Output (₹ / 1M tokens)
Qwen3 1.7B	0	0
Qwen3 235B-A22B	9.07	55.44
Qwen3.8-27B	16.32	48.96

Usage limits may apply to keep the free tier sustainable for everyone — check /models for the current terms on the model page before relying on it for production volume.

Calling Qwen3 1.7B from the unoblox endpoint

curl https://api.unoblox.ai/v1/chat/completions \
  -H "Authorization: Bearer ub-gw-xxxxxxxxxxxxxxxx" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "qwen/qwen3-1.7b",
    "messages": [{"role": "user", "content": "Rewrite this sentence in plain English."}]
  }'

The same call shape works for every other model on unoblox — once an integration is validated against Qwen3 1.7B, pointing it at a larger model for production is a one-line change to the model field.

What to build on a free model

• A first integration test — confirm auth, request shape and streaming before writing a line of billing-aware code.
• Lightweight, low-stakes tasks — short rewrites, simple classification, draft generation where a small model's quality is enough.
• A fallback tier — some teams route a request to Qwen3 1.7B first and only escalate to a paid model when the small model's answer isn't good enough.

For anything customer-facing at real volume, it's worth comparing Qwen3 1.7B's output quality against a paid small model like Qwen3 235B-A22B or GPT-5 mini before committing.

Frequently asked questions

Is Qwen3 1.7B really free on unoblox? Yes, it's priced at ₹0 per token, unlike every other model in the catalogue.

Do I need a payment method to call Qwen3 1.7B? No card is required to start; the free model is usable as soon as an account and API key exist.

Are there usage limits on the free tier? Limits may apply to keep the tier sustainable — check the model page for current terms before planning production volume around it.

Is Qwen3 1.7B good enough for production traffic? It depends on the task. It's a small model, well suited to lightweight or low-stakes work; test it against a paid model before relying on it for anything customer-critical.

Does the free model use the same API key as paid models? Yes, the same ub-gw- key and base URL work across Qwen3 1.7B and every paid model.

Can I switch from Qwen3 1.7B to a paid model without rewriting my integration? Yes, only the model field changes — the request and response shape stay the same.

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