Mistral AI LLM Provider

The Mistral provider connects Beluga AI to Mistral AI’s models using the mistral-go SDK. It supports chat completions, streaming, tool calling, and JSON mode.

Choose Mistral when you need competitive model quality with European data residency (hosted in EU). Mistral Large offers strong reasoning and multilingual capabilities, while Codestral specializes in code generation. Mistral also offers efficient open-weight models that can be self-hosted.

Installation

go get github.com/lookatitude/beluga-ai/llm/providers/mistral

Configuration

Field	Required	Default	Description
Model	No	"mistral-large-latest"	Model ID
APIKey	Yes	—	Mistral API key
BaseURL	No	https://api.mistral.ai	Override API endpoint
Timeout	No	30s	Request timeout

Environment variables:

Variable	Maps to
MISTRAL_API_KEY	APIKey

Basic Usage

package main

import (
    "context"
    "fmt"
    "log"
    "os"

    "github.com/lookatitude/beluga-ai/config"
    "github.com/lookatitude/beluga-ai/llm"
    "github.com/lookatitude/beluga-ai/schema"
    _ "github.com/lookatitude/beluga-ai/llm/providers/mistral"
)

func main() {
    model, err := llm.New("mistral", config.ProviderConfig{
        Model:  "mistral-large-latest",
        APIKey: os.Getenv("MISTRAL_API_KEY"),
    })
    if err != nil {
        log.Fatal(err)
    }

    msgs := []schema.Message{
        schema.NewSystemMessage("You are a helpful assistant."),
        schema.NewHumanMessage("What is the capital of France?"),
    }

    resp, err := model.Generate(context.Background(), msgs)
    if err != nil {
        log.Fatal(err)
    }

    fmt.Println(resp.Text())
}

Streaming

for chunk, err := range model.Stream(context.Background(), msgs) {
    if err != nil {
        log.Fatal(err)
    }
    fmt.Print(chunk.Delta)
}
fmt.Println()

The Mistral streaming implementation respects context cancellation and returns channel-based events converted to iter.Seq2.

Advanced Features

Tool Calling

tools := []schema.ToolDefinition{
    {
        Name:        "get_weather",
        Description: "Get current weather for a location",
        InputSchema: map[string]any{
            "type": "object",
            "properties": map[string]any{
                "location": map[string]any{
                    "type":        "string",
                    "description": "City name",
                },
            },
            "required": []any{"location"},
        },
    },
}

modelWithTools := model.BindTools(tools)
resp, err := modelWithTools.Generate(ctx, msgs, llm.WithToolChoice(llm.ToolChoiceAuto))
if err != nil {
    log.Fatal(err)
}

for _, tc := range resp.ToolCalls {
    fmt.Printf("Tool: %s, Args: %s\n", tc.Name, tc.Arguments)
}

Mistral tool choice mapping:

Beluga ToolChoice	Mistral Equivalent
llm.ToolChoiceAuto	auto
llm.ToolChoiceNone	none
llm.ToolChoiceRequired	any

JSON Mode

resp, err := model.Generate(ctx, msgs,
    llm.WithResponseFormat(llm.ResponseFormat{Type: "json_object"}),
)

Generation Options

The provider defaults to Temperature: 0.7 and TopP: 1.0 unless overridden:

resp, err := model.Generate(ctx, msgs,
    llm.WithTemperature(0.3),
    llm.WithMaxTokens(2048),
    llm.WithTopP(0.9),
)

Error Handling

resp, err := model.Generate(ctx, msgs)
if err != nil {
    // Errors are wrapped with the "mistral:" prefix
    log.Fatal(err)
}

Direct Construction

import "github.com/lookatitude/beluga-ai/llm/providers/mistral"

model, err := mistral.New(config.ProviderConfig{
    Model:  "mistral-large-latest",
    APIKey: os.Getenv("MISTRAL_API_KEY"),
})

Available Models

Model ID	Description
mistral-large-latest	Most capable Mistral model
mistral-small-latest	Fast, efficient model
codestral-latest	Code generation specialist
open-mistral-nemo	Open-weight 12B model

Refer to Mistral AI’s documentation for the latest model list.