Kimi K3.1 Pro API quickstart: model ID, base_url, and OpenAI SDK calls

One-sentence takeaway

To integrate Kimi K3.1 Pro, use the model ID kimi-k3.1-pro, base_url=https://api.moonshot.ai/v1, and the OpenAI SDK's Chat Completions interface. This flagship model is unlocked only after a successful top-up of at least $1; the cumulative top-up amount determines the account tier and the concurrency, RPM, TPM, and TPD limits.

Use cases

• Suitable tasks: Integrating Kimi K3.1 Pro into a Python service, script, or client that supports OpenAI-compatible APIs, and checking the account unlock requirements.

• Unsuitable tasks: Reasoning levels, vision input, context caching, tool calling, or model performance evaluation; these topics are outside this article's scope.

• Applicable model version: The flagship model kimi-k3.1-pro; the official page also lists the high-speed variant kimi-k3.1-pro-highspeed, whose pricing and availability require separate verification.

• Applicable clients, agents, or APIs: Python 3.9+, openai>=1.0, Kimi API Chat Completions, and clients using the same OpenAI-compatible endpoint.

• Recommended reasoning level and parameters: This article does not cover reasoning configuration. If you need to set reasoning_effort, consult the official Reasoning Effort documentation.

Ready-to-use content

First install the OpenAI Python SDK and put the API key in an environment variable:

python3 -m pip install --upgrade 'openai>=1.0'
export MOONSHOT_API_KEY='YOUR_API_KEY'

Minimal runnable call:

import os

from openai import OpenAI

client = OpenAI(
    api_key=os.environ["MOONSHOT_API_KEY"],
    base_url="https://api.moonshot.ai/v1",
)

completion = client.chat.completions.create(
    model="kimi-k3.1-pro",
    messages=[
        {"role": "user", "content": "Introduce Kimi K3.1 Pro in one sentence."}
    ],
)

print(completion.choices[0].message.content)

Configuration mapping:

Testing or workflow steps

1. Create an API key on Kimi API Platform and write it to MOONSHOT_API_KEY; do not put the key in the code repository.

2. Confirm that the account has completed a successful top-up of at least $1 so that Kimi K3.1 Pro is unlocked.

3. Install Python 3.9+ and openai>=1.0, then use the example above with the specified base_url and model ID.

4. Send a minimal Chat Completions request and check whether completion.choices[0].message.content returns content.

5. Before going live, check the concurrency, RPM, TPM, and TPD limits for the account tier associated with the cumulative top-up amount, and record request failures and rate-limit responses in your service.

Original evidence and data

• The official page's Get started section requires Python 3.9+ and the OpenAI SDK, and provides python3 -m pip install --upgrade 'openai>=1.0'.

• The official initialization example uses api_key=os.environ["MOONSHOT_API_KEY"] and base_url="https://api.moonshot.ai/v1".

• The official Basic call example uses client.chat.completions.create, with "kimi-k3.1-pro" in the model field; the page also lists the high-speed variant kimi-k3.1-pro-highspeed.

• The official Access requirements section states that Kimi K3.1 Pro is the flagship model and that the minimum successful top-up amount is $1; the cumulative top-up amount determines the account tier and rate limits, including concurrency, RPM, TPM, and TPD, and links to Recharge and Rate Limits.

• The page's structured data has dateModified set to 2026-10-23T18:17:30.027Z, but the page has no publication date field, so the publication date is recorded as “Not specified.”

• The code blocks on the page are official configuration examples. This article did not use an API key to make a local request, so it contains no locally measured latency, success rate, or response-content data.

Scope and limitations

• $1 is the minimum top-up amount given by the official page for unlocking the flagship model. It is not the model call price and does not mean that every account receives the same concurrency or rate limits.

• The account tier and rate-limit values are determined by cumulative top-ups and platform rules. Actual limits should be checked in the account console and in the current content of Recharge and Rate Limits.

• kimi-k3.1-pro-highspeed is another model ID mentioned on the official page. This article does not extrapolate its speed, price, or performance to the flagship model.

• This article records only the basic API configuration. It does not assess model capability, treat unfilled benchmark placeholders such as [xx.x%] on the official page as numbers, or replace dedicated articles on reasoning, vision, caching, and tool calling.

Reproduction notes

1. Open the original article and confirm that the page title is Kimi K3.1 Pro, then record the collection date, model ID, Base URL, SDK version requirement, and top-up limitation.

2. In a local Python 3.9+ environment, install openai>=1.0, set MOONSHOT_API_KEY, and run the script in “Ready-to-use content.”

3. Before using a real API key, confirm that the account has completed a successful top-up of at least $1; then record the HTTP status, response content, model ID, and rate-limit information.

4. This reproduction only verifies the configuration on the official page. No real API request was submitted, so it cannot claim to have local call results.