X.AI (Grok)¶

Overview¶

Models: Grok-4.1-Fast, Grok-4, Grok-4-Fast, Grok-Code-Fast, Grok-3, Grok-3-Mini, Grok-2, Grok-2-Vision
Features: Chat completions, streaming, OpenAI-compatible API, 2M context window

Configuration¶

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

Available Models¶

Model	Context Window	Description
`omnillm.ModelGrok41Fast`	2M	Grok 4.1 Fast (reasoning/non-reasoning)
`omnillm.ModelGrok4`	128K	Grok 4 (0709)
`omnillm.ModelGrok4Fast`	2M	Grok 4 Fast (reasoning/non-reasoning)
`omnillm.ModelGrokCodeFast`	128K	Grok Code Fast
`omnillm.ModelGrok3`	128K	Grok 3
`omnillm.ModelGrok3Mini`	128K	Grok 3 Mini
`omnillm.ModelGrok2`	128K	Grok 2
`omnillm.ModelGrok2Vision`	128K	Grok 2 Vision

OpenAI Compatibility¶

X.AI uses an OpenAI-compatible API, so parameters like Seed, PresencePenalty, and FrequencyPenalty are supported:

response, err := client.CreateChatCompletion(ctx, &omnillm.ChatCompletionRequest{
    Model: omnillm.ModelGrok4Fast,
    Messages: messages,
    Seed: &[]int{42}[0],
    PresencePenalty: &[]float64{0.5}[0],
})

Reasoning¶

X.AI Grok models support the reasoning_effort parameter (OpenAI-compatible):

effort := omnillm.ReasoningEffortHigh
response, err := client.CreateChatCompletion(ctx, &omnillm.ChatCompletionRequest{
    Model:           omnillm.ModelGrok4Fast,
    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¶

grok-4.1-fast - Supports reasoning mode
grok-4-fast - Supports reasoning mode

See Reasoning Feature Guide for cross-provider usage.