Kimi K3.1 Pro API Pricing in TWD: AA Index 55.1, $2 / $10, 1M Context
Kimi K3.1 Pro (Moonshot AI) is live on BazaarLink at $2 input / $10 output per million tokens with a 1M-token context. Artificial Analysis Intelligence Index 55.1, up from Kimi K3's 43.6 at a lower price.
Query date: 2026-10-24 (Taipei time). Benchmark data is from the Artificial Analysis (AA) data feed. Prices are the USD unit prices on the BazaarLink models page; the exchange rate is 1 USD = NT$ 31.5.
Kimi K3.1 Pro is Moonshot AI's latest flagship, released on 2026-10-23. It is now available on BazaarLink as kimi-k3.1-pro.
Bottom line up front: K3.1 Pro scores 55.1 on the Artificial Analysis Intelligence Index, against 43.6 for Kimi K3, while its price drops from $3 / $15 to $2 / $10 per million tokens. The context window stays at 1M tokens.
Price (USD ↔ NT$)
| Item | USD per MTok | NT$ per MTok |
|---|---|---|
| Input | $2.00 | NT$ 63 |
| Output | $10.00 | NT$ 315 |
| Cache hit (read) | $0.20 | NT$ 6.3 |
| Cache write | $2.50 | NT$ 78.75 |
| Model | Input | Output | AA Intelligence Index |
|---|---|---|---|
| Kimi K3.1 Pro | $2.00 | $10.00 | 55.1 |
| Kimi K3 (Max) | $3.00 | $15.00 | 43.6 |
Source for the index: Artificial Analysis, Kimi K3.1 Pro (query date 2026-10-24). For current unit prices see BazaarLink Models.
Specifications
- Context length: 1,048,576 tokens
- Input: text, image, video per the model catalog → Output: text
- Supports: reasoning, tool calling (tools), structured outputs
- Model ID:
kimi-k3.1-pro(aliasmoonshotai/kimi-k3.1-pro)
How to call it
curl https://api.bazaarlink.ai/v1/chat/completions \
-H "Authorization: Bearer sk-bl-YOUR_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "kimi-k3.1-pro",
"messages": [{"role": "user", "content": "Hello"}]
}'
https://bazaarlink.ai/api/v1 works the same way. It is OpenAI-compatible: change base_url and model and nothing else.
Cost estimate: coding assistant (NT$/month)
Assumptions: 200 requests per day, each with 8,000 input tokens and 1,500 output tokens, no cache hits.
- Monthly requests: 6,000; input 48 MTok, output 9 MTok
- Input: 48 × NT$ 63 = NT$ 3,024
- Output: 9 × NT$ 315 = NT$ 2,835
- Total: about NT$ 5,859 per month (≈ USD 186). With the same traffic on Kimi K3 it would be about NT$ 8,789.
Cache hits cost a tenth of normal input, so workloads that reuse a long system prompt or repository context get cheaper still.
Who should use it
Suitable for: coding and agent workloads that want a higher AA score than Kimi K3 at a lower price; long-context work up to 1M tokens.
Think twice: reasoning models spend extra output tokens on thinking, so budget max_tokens generously (a few hundred tokens at minimum) and estimate cost by output length, not only by unit price.
Want TWD billing and a Taiwan unified invoice for this model? See the guide to expense reporting for AI APIs in Taiwan, or try BazaarLink.
Explore current model discounts and learn how usage rebates add credit to your balance.
View discounts and usage rebatesFAQ
What is the model ID for Kimi K3.1 Pro?
The model ID is kimi-k3.1-pro, with the compatible alias moonshotai/kimi-k3.1-pro. Either value works for the model parameter, and the endpoint is OpenAI-compatible, so only base_url and model need to change.
How much does Kimi K3.1 Pro cost?
US$2.00 per million input tokens and US$10.00 per million output tokens. Cache-hit input is US$0.20 per million tokens and cache writes are US$2.50 per million tokens. At 1 USD = NT$31.5 that is about NT$63 input and NT$315 output per million tokens.
Is Kimi K3.1 Pro cheaper than Kimi K3?
Yes. Kimi K3 is US$3.00 input / US$15.00 output, so K3.1 Pro is one third cheaper on both, and scores higher on the Artificial Analysis Intelligence Index (55.1 vs 43.6).
Does Kimi K3.1 Pro support tool calling and reasoning?
Yes. We called it with tools and tool_choice and got a tool call back, and the responses report reasoning tokens in usage. The context length is 1,048,576 tokens.
TWD billing · Taiwan invoices · leading AI models · OpenAI-compatible API