Bridge by Wekoya / Models / Moonshot AI / Kimi K2 Thinking API, priced in naira

Kimi K2 Thinking API, priced in naira

In short: You can use Kimi K2 Thinking, a model by Moonshot AI, from Nigeria through Bridge and pay in naira, with no dollar card. It costs ₦1,053 per million input tokens and ₦4,388 per million output tokens. Call it with the OpenAI SDK using the model ID kimi-k2-thinking.

Updated

Price and limits

Kimi K2 Thinking
Input, per million tokens₦1,053
Output, per million tokens₦4,388
Cached input, per million tokens₦1,053
Writing to the cache, per million tokens₦1,053
Context window262,144 tokens
Longest reply16,000 tokens
Reads picturesNo
Tool callingYes
Reasoning (thinks before it answers)Yes
Model IDkimi-k2-thinking

Prices as of 8 November 2026. They are in naira and follow the exchange rate, so the live price may differ a little. A token is roughly three quarters of a word.

What a request costs

A request that sends 1,500 tokens (instructions, the conversation so far and a question) and gets 500 back (about 375 words) costs about ₦3.77 on 8 November 2026. A thousand of them cost about ₦3,773.

Use Kimi K2 Thinking from your code

Bridge speaks the OpenAI format, so the OpenAI SDK works as it is: set the base URL to https://wekoya-api.wekoya.tech/v1, use your Bridge key, and pass kimi-k2-thinking as the model. Create a key, then:

from openai import OpenAI

client = OpenAI(api_key="wk_live_...", base_url="https://wekoya-api.wekoya.tech/v1")
reply = client.chat.completions.create(
    model="kimi-k2-thinking",
    messages=[{"role": "user", "content": "Hello"}],
)
print(reply.choices[0].message.content)

Streaming works the same way, and so does tool calling. The quickstart has the same in JavaScript and curl, and the Responses API works too.

Other Moonshot AI models on Bridge

ModelInput / 1MOutput / 1MContext
Kimi K2.5₦1,053₦5,265262,144
Kimi K2.7 Code₦1,193₦5,967262,144
Kimi K2.6₦1,316₦6,143262,144
Kimi K3₦5,002₦25,0091,048,576

Every Moonshot AI model on Bridge, with what each can do.

Lower-priced models that can do the same things

These cost less per request than Kimi K2 Thinking and can also call tools, reason and take at least 262,144 tokens of context. Quality differs from model to model, so test on your own examples first: AI tasks compares models on your cases and tells you whether a cheaper one is good enough.

ModelInput / 1MOutput / 1MContext
GLM-5.3₦988₦4,3881,048,576
GLM-5.2₦986₦3,1591,048,576
Nova 2 Lite₦527₦4,3881,000,000
Gemini 3.5 Flash Lite₦446₦3,7131,048,576

Create a free account

Common questions

How much does Kimi K2 Thinking cost in naira?

On Bridge, Kimi K2 Thinking costs ₦1,053 per million input tokens and ₦4,388 per million output tokens, with cached input at ₦1,053 (as of 8 November 2026). A request that sends 1,500 tokens and gets 500 back costs about ₦3.77. There is no subscription and no minimum.

Can I use the Kimi K2 Thinking API from Nigeria without a dollar card?

Yes. Create a free Bridge account, top up your naira balance by bank transfer to your own account number, and call Kimi K2 Thinking with your Bridge key. A card works too, but it is not needed.

Does Kimi K2 Thinking on Bridge work with the OpenAI SDK?

Yes. Use the OpenAI SDK or any OpenAI-compatible client with the base URL https://wekoya-api.wekoya.tech/v1 and the model ID kimi-k2-thinking. Streaming works, and so does tool calling.

What is the context window of Kimi K2 Thinking?

262,144 tokens, about 196,608 words of instructions, conversation and documents together. It writes up to 16,000 tokens in one reply.

Can Kimi K2 Thinking read pictures?

No. Kimi K2 Thinking reads text only. For pictures, pick a model marked "Reads pictures" on the models page.

Read next