Kimi LLM | LiveKit Docs

Overview

LiveKit Inference offers Kimi models through Baseten. Pricing is available on the pricing page.

Model name	Model ID	Providers
Kimi K2 Instruct	moonshotai/kimi-k2-instruct	baseten

Usage

To use Kimi, pass the model id to the llm argument in your AgentSession. LiveKit Inference manages the connection to the best available provider automatically.

from livekit.agents import AgentSession

session = AgentSession(
    llm="moonshotai/kimi-k2-instruct",
    # ... tts, stt, vad, turn_detection, etc.
)

import { AgentSession } from '@livekit/agents';

session = new AgentSession({
    llm: "moonshotai/kimi-k2-instruct",
    // ... tts, stt, vad, turn_detection, etc.
});

Parameters

To customize additional parameters, including the specific provider to use, use the LLM class from the inference module.

from livekit.agents import AgentSession, inference

session = AgentSession(
    llm=inference.LLM(
        model="moonshotai/kimi-k2-instruct", 
        provider="baseten",
        extra_kwargs={
            "max_completion_tokens": 1000
        }
    ),
    # ... tts, stt, vad, turn_detection, etc.
)

import { AgentSession, inference } from '@livekit/agents';

session = new AgentSession({
    llm: new inference.LLM({ 
        model: "moonshotai/kimi-k2-instruct", 
        provider: "baseten",
        extraKwargs: { 
            max_completion_tokens: 1000 
        }
    }),
    // ... tts, stt, vad, turn_detection, etc.
});

modelstringRequired

The model ID from the models list.

providerstringOptional

Set a specific provider to use for the LLM. Refer to the models list for available providers. If not set, LiveKit Inference uses the best available provider, and bills accordingly.

extra_kwargsdictOptional

Additional parameters to pass to the provider's Chat Completions API, such as max_completion_tokens. See the provider's documentation for more information.

Additional resources

The following links provide more information about Kimi in LiveKit Inference.

Baseten Plugin

Plugin to use your own Baseten account instead of LiveKit Inference.

Baseten docs

Baseten's official Model API documentation.