Skip to content

OmniLLM Core

Kimi (Moonshot AI)

plexusone/omnillm-core

Kimi (Moonshot AI)¶

Overview¶

Models: Kimi K2.5, Kimi K2 series, Moonshot V1 series
Features: Chat completions, streaming, long context (up to 256K), multimodal support

Configuration¶

client, err := omnillm.NewClient(omnillm.ClientConfig{
    Providers: []omnillm.ProviderConfig{
        {Provider: omnillm.ProviderNameKimi, APIKey: "your-kimi-api-key"},
    },
})

Available Models¶

Kimi K2.5 Series¶

Model	Context Window	Description
`kimi-k2.5`	256K	Most intelligent multimodal model

Kimi K2 Series¶

Model	Context Window	Description
`kimi-k2-0905-preview`	256K	Enhanced agentic coding model
`kimi-k2-0711-preview`	128K	MoE base model with code/agent capabilities
`kimi-k2-turbo-preview`	256K	High-speed version (60-100 tokens/sec)
`kimi-k2-thinking`	256K	Long-term thinking with multi-step tool usage
`kimi-k2-thinking-turbo`	256K	High-speed thinking model

Moonshot V1 Series¶

Model	Context Window	Description
`moonshot-v1-8k`	8K	Short text generation
`moonshot-v1-32k`	32K	Long text generation
`moonshot-v1-128k`	128K	Very long text generation
`moonshot-v1-*-vision-preview`	8K/32K/128K	Image understanding

OpenAI Compatibility¶

Kimi uses an OpenAI-compatible API:

response, err := client.CreateChatCompletion(ctx, &omnillm.ChatCompletionRequest{
    Model: "moonshot-v1-8k",
    Messages: []omnillm.Message{
        {Role: omnillm.RoleUser, Content: "Hello!"},
    },
    Temperature: &[]float64{0.7}[0],
    MaxTokens:   &[]int{1000}[0],
})

Streaming¶

stream, err := client.CreateChatCompletionStream(ctx, &omnillm.ChatCompletionRequest{
    Model: "kimi-k2.5",
    Messages: messages,
})
if err != nil {
    log.Fatal(err)
}
defer stream.Close()

for {
    chunk, err := stream.Recv()
    if err == io.EOF {
        break
    }
    if err != nil {
        log.Fatal(err)
    }
    fmt.Print(chunk.Choices[0].Delta.Content)
}

Reasoning¶

Kimi models support the reasoning_effort parameter (OpenAI-compatible):

effort := omnillm.ReasoningEffortHigh
response, err := client.CreateChatCompletion(ctx, &omnillm.ChatCompletionRequest{
    Model:           "kimi-k2-thinking",
    ReasoningEffort: &effort,
    Messages: []omnillm.Message{
        {Role: omnillm.RoleUser, Content: "Solve this complex problem..."},
    },
})

ReasoningEffort Values¶

Constant	Value	Description
`omnillm.ReasoningEffortNone`	`"none"`	Disable reasoning
`omnillm.ReasoningEffortLow`	`"low"`	Light reasoning
`omnillm.ReasoningEffortMedium`	`"medium"`	Moderate reasoning
`omnillm.ReasoningEffortHigh`	`"high"`	Deep reasoning

Reasoning-Capable Models¶

kimi-k2-thinking - Long-term thinking with multi-step tool usage
kimi-k2-thinking-turbo - High-speed thinking model

See Reasoning Feature Guide for cross-provider usage.

API Documentation¶