Kimi K3 pricing in India (₹)

Moonshot AI's Kimi K3 (moonshotai/kimi-k3) is billed in rupees on unoblox. This page sets out the rates and shows what typical volumes cost. It also puts K3 beside a few other models in the catalog, all of them computed from live rates.

Kimi K3 rates

Rates below are as listed on 7 October 2026; the model page always shows the live rate.

Price leg	₹ per 1M tokens
Input	₹288.61
Cached input	₹28.86
Cache write	₹360.76
Output	₹1,443.08

Every listed rate already includes unoblox's 5% platform fee, and you are charged exactly the listed rate. Cached input is charged on prompt tokens the provider serves from its prompt cache. Prompt tokens written to the cache are charged at the cache-write rate (₹360.76), and all remaining prompt tokens at the input rate. Reasoning tokens are billed as output. The full formula is in Pricing & credits.

What volumes cost

Each figure below is input tokens × input rate + output tokens × output rate, with no caching:

Monthly volume	Kimi K3	Claude Sonnet 5	DeepSeek V4 Pro (0813)
1 M in + 0.25 M out	₹649.38	₹455.70	₹197.46
10 M in + 2 M out	₹5,772.26	₹4,050.68	₹1,842.98
100 M in + 20 M out	₹57,722.60	₹40,506.80	₹18,429.80

Per million tokens, Kimi K3's input rate is 143% of Claude Sonnet 5's (₹202.53) and 57% of Claude Opus 5's (₹506.35). Its output rate is 142% of Sonnet 5's (₹1,012.69) and 57% of Opus 5's (₹2,531.73). Price is only part of the decision, so test each model on your own prompts.

How caching changes the bill

Take a chatbot that sends 10 M input tokens a month. A long system prompt is reused, so 60% of those tokens (6 M) are served from cache. It also produces 2 M output tokens:

• Uncached input: 4 × ₹288.61 = ₹1,154.44
• Cached input: 6 × ₹28.86 = ₹173.16
• Output: 2 × ₹1,443.08 = ₹2,886.16
• Total: ₹4,213.76, against ₹5,772.26 with no cache hits

This example assumes no tokens are billed as cache writes; any the provider reports as written to the cache cost ₹360.76 per 1M. How much of a prompt the provider serves from cache depends on the provider. The share it reports for each request is what gets billed.

Calling Kimi K3

from openai import OpenAI
client = OpenAI(base_url="https://api.unoblox.ai/v1", api_key="ub-gw-…")
response = client.chat.completions.create(
    model="moonshotai/kimi-k3",
    messages=[{"role": "user", "content": "Solve for x: 2x + 5 = 13"}],
)
print(response.usage)

The response returns token usage. To see the rupee cost, check the usage view in your dashboard.

Billing

Usage is paid from a prepaid rupee wallet. GST at 18% is added when you top up (pay ₹118 and ₹100 of credit lands in your wallet), and one GST tax invoice is issued for each top-up; individual requests carry no separate invoice. Credits do not expire.

Frequently asked questions

Does Kimi K3 accept images? Yes. The catalog lists text and image input. Its context window is 1,048,576 tokens.

Is GST included in these rates? No. Per-token rates exclude GST. GST is added once, when you top up.

Can I cap what a key spends? Yes. Set a per-key spend cap or a workspace budget. A request that would breach a hard cap is refused instead of overspending.

Is there a free model I can test with first? Yes. Qwen3.8-27B is priced at ₹0. Switching to Kimi K3 later only changes the model field.

Where is the live rate? On the Kimi K3 model page.

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