OpenLLMetry is a set of extensions built on top of OpenTelemetry that gives you complete observability over your LLM application. Because it uses OpenTelemetry under the hood, it can be connected to your existing observability solutions - Datadog, Honeycomb, and others.

The repo contains standard OpenTelemetry instrumentations for LLM providers and Vector DBs, as well as a Traceloop SDK that makes it easy to get started with OpenLLMetry, while still outputting standard OpenTelemetry data that can be connected to your observability stack. If you already have OpenTelemetry instrumented, you can just add any of our instrumentations directly.

The easiest way to get started is to use our SDK. For a complete guide, go to our docs.

go get github.com/traceloop/go-openllmetry/traceloop-sdk

package main

import (
	"context"

	sdk "github.com/traceloop/go-openllmetry/traceloop-sdk"
)

func main() {
    ctx := context.Background()

    traceloop := sdk.NewClient(ctx, sdk.Config{
		APIKey: os.Getenv("TRACELOOP_API_KEY"),
	})
	defer func() { traceloop.Shutdown(ctx) }()
}

OpenLLMetry is in early-alpha exploratory stage, and we're still figuring out what to instrument. As opposed to other languages, there aren't many official LLM libraries (yet?), so for now you'll have to manually log prompts:

package main

import (
	"context"
	"fmt"
	"os"

	"github.com/sashabaranov/go-openai"
	sdk "github.com/traceloop/go-openllmetry/traceloop-sdk"
)

func main() {
	ctx := context.Background()

	// Initialize Traceloop
	traceloop := sdk.NewClient(ctx, config.Config{
		APIKey:  os.Getenv("TRACELOOP_API_KEY"),
	})
	defer func() { traceloop.Shutdown(ctx) }()

	// Call OpenAI like you normally would
	resp, err := client.CreateChatCompletion(
		context.Background(),
		openai.ChatCompletionRequest{
			Model: openai.GPT3Dot5Turbo,
			Messages: []openai.ChatCompletionMessage{
				{
					Role:    openai.ChatMessageRoleUser,
					Content: "Tell me a joke about OpenTelemetry!",
				},
			},
		},
	)

    var promptMsgs []sdk.Message
    for i, message := range request.Messages {
    	promptMsgs = append(promptMsgs, sdk.Message{
    		Index:   i,
    		Content: message.Content,
    		Role:    message.Role,
    	})
    }

	// Log the request
    llmSpan, err := traceloop.LogPrompt(
    	ctx,
    	sdk.Prompt{
    		Vendor: "openai",
    		Mode:   "chat",
    		Model: request.Model,
    		Messages: promptMsgs,
    	},
    	sdk.TraceloopAttributes{
    		WorkflowName: "example-workflow",
    		EntityName:   "example-entity",
    	},
    )
    if err != nil {
    	fmt.Printf("LogPrompt error: %v\n", err)
    	return
    }

    client := openai.NewClient(os.Getenv("OPENAI_API_KEY"))
    resp, err := client.CreateChatCompletion(
    	context.Background(),
    	*request,
    )
    if err != nil {
    	fmt.Printf("ChatCompletion error: %v\n", err)
    	return
    }

    var completionMsgs []sdk.Message
    for _, choice := range resp.Choices {
    	completionMsgs = append(completionMsgs, sdk.Message{
    		Index:   choice.Index,
    		Content: choice.Message.Content,
    		Role:    choice.Message.Role,
    	})
    }

	// Log the response
    llmSpan.LogCompletion(ctx, sdk.Completion{
    	Model:    resp.Model,
    	Messages: completionMsgs,
    }, sdk.Usage{
    	TotalTokens:       resp.Usage.TotalTokens,
    	CompletionTokens:  resp.Usage.CompletionTokens,
    	PromptTokens:      resp.Usage.PromptTokens,
    })
}

Whether it's big or small, we love contributions ❤️ Check out our guide to see how to get started.

For Go

Get started »

Slack | Docs | Website

🚀 Getting Started

⏫ Supported (and tested) destinations

🪗 What do we instrument?

🌱 Contributing

💚 Community & Support

🙏 Special Thanks