LLM Inference · Qwen3.6-27B
Inference from ₹20 a million input tokens.
Qwen/Qwen3.6-27B:excloud is a 27B parameter general-purpose model. Input and output tokens are priced separately and billed per million tokens. There is no minimum spend — use it once, pay for what you used.
per million tokens, each way
- 1M input tokens × ₹20/1M tokens
- ₹20
- 1M output tokens × ₹60/1M tokens
- ₹60
Rates
Input and output, billed separately.
Input tokens cover the prompt and any context you send; output tokens cover what the model generates back. Both are billed per million tokens. A single request is charged for exactly the tokens it consumed — no minimum spend required.
| Item | What it covers | Rate |
|---|---|---|
| Input tokens | prompt + any context | ₹20/1M tokens |
| Output tokens | the generated response | ₹60/1M tokens |
| Minimum commit | — | none |
The model
One model, 27 billion parameters.
Qwen3.6-27B handles text and code generation, question answering over context you provide, and work across languages. The rate is the same regardless of what you ask it to do; the bill tracks only tokens in and tokens out.
Get started
The math is right there in the response.
A short prompt and reply costs a fraction of a rupee. The API returns the token counts with every response, so you can verify the bill yourself from the first request.