SambaNova LLM Provider

The SambaNova provider connects Beluga AI to SambaNova’s inference platform, which provides high-throughput inference on custom RDU (Reconfigurable Dataflow Unit) hardware. SambaNova exposes an OpenAI-compatible API, so this provider supports all standard features including streaming, tool calling, and structured output.

Choose SambaNova when you need high-throughput batch inference or consistent low-latency serving of large open-source models. SambaNova’s custom RDU hardware is optimized for sustained throughput, making it well-suited for production workloads with predictable high-volume traffic.

Installation

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

Configuration

Field	Required	Default	Description
Model	Yes	—	Model ID (e.g. "Meta-Llama-3.3-70B-Instruct")
APIKey	Yes	—	SambaNova API key (sn-...)
BaseURL	No	https://api.sambanova.ai/v1	Override API endpoint
Timeout	No	30s	Request timeout

Environment variables:

Variable	Maps to
SAMBANOVA_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/sambanova"
)

func main() {
    model, err := llm.New("sambanova", config.ProviderConfig{
        Model:  "Meta-Llama-3.3-70B-Instruct",
        APIKey: os.Getenv("SAMBANOVA_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()

Advanced Features

Tool Calling

modelWithTools := model.BindTools(tools)
resp, err := modelWithTools.Generate(ctx, msgs, llm.WithToolChoice(llm.ToolChoiceAuto))

Structured Output

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

Generation Options

resp, err := model.Generate(ctx, msgs,
    llm.WithTemperature(0.7),
    llm.WithMaxTokens(2048),
    llm.WithTopP(0.9),
    llm.WithStopSequences("END"),
)

Error Handling

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

Direct Construction

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

model, err := sambanova.New(config.ProviderConfig{
    Model:  "Meta-Llama-3.3-70B-Instruct",
    APIKey: os.Getenv("SAMBANOVA_API_KEY"),
})

Available Models

Model ID	Description
Meta-Llama-3.3-70B-Instruct	Llama 3.3 70B — best quality/speed tradeoff
Meta-Llama-3.1-8B-Instruct	Llama 3.1 8B — fastest, lowest cost
Meta-Llama-3.1-405B-Instruct	Llama 3.1 405B — most capable, highest throughput

Refer to SambaNova’s documentation for the latest model list.