# OpenLIT — OpenTelemetry LLM Observability

> Monitor LLM costs, latency, and quality with OpenTelemetry-native tracing. GPU monitoring and guardrails built in. 2.3K+ stars.

## Install

Save in your project root:

# OpenLIT — OpenTelemetry LLM Observability

## Quick Use

```bash
pip install openlit
```

```python
import openlit

# One line to instrument all LLM calls
openlit.init()

# Now use any LLM library as usual — OpenLIT traces automatically
from openai import OpenAI
client = OpenAI()

response = client.chat.completions.create(
    model="gpt-4o",
    messages=[{"role": "user", "content": "What is RAG?"}]
)
# Traces, costs, latency, and token usage are captured automatically
```

Launch the dashboard:
```bash
docker run -p 3000:3000 ghcr.io/openlit/openlit
```

---

## Intro

OpenLIT is an OpenTelemetry-native observability platform with 2,300+ GitHub stars purpose-built for LLM applications. With a single `openlit.init()` call, it auto-instruments 50+ LLM providers and frameworks, capturing traces, cost tracking, token usage, latency metrics, and quality evaluations. It includes GPU monitoring, prompt/response guardrails, and a built-in dashboard — all built on open standards (OpenTelemetry) so your data isn't locked into a proprietary vendor. Send traces to any OTLP-compatible backend (Grafana, Datadog, Jaeger).

Works with: OpenAI, Anthropic, Google, Ollama, LangChain, LlamaIndex, CrewAI, Haystack, and 50+ more. Best for teams running LLM apps in production who need cost and quality monitoring. Setup time: under 2 minutes.

---

## OpenLIT Features

### One-Line Auto-Instrumentation

```python
import openlit
openlit.init()  # That's it!

# All these are now automatically traced:
# - OpenAI calls
# - Anthropic calls
# - LangChain chains
# - LlamaIndex queries
# - Vector DB operations
# - Embedding generation
```

### Supported Providers (50+)

| Category | Providers |
|----------|----------|
| **LLMs** | OpenAI, Anthropic, Google, Cohere, Mistral, Ollama, HuggingFace |
| **Frameworks** | LangChain, LlamaIndex, CrewAI, Haystack, AutoGen |
| **Vector DBs** | Chroma, Pinecone, Qdrant, Weaviate, Milvus |
| **Embeddings** | OpenAI, Cohere, HuggingFace, Voyage AI |
| **GPUs** | NVIDIA (via nvidia-smi monitoring) |

### Cost Tracking

```python
# OpenLIT automatically calculates costs per request
# Dashboard shows:
# - Total spend per model
# - Cost per request
# - Token usage breakdown (input vs output)
# - Cost trends over time
# - Budget alerts
```

### Guardrails

```python
import openlit

# Add safety guardrails
openlit.init(
    guardrails=[
        openlit.guardrails.ToxicityGuardrail(threshold=0.8),
        openlit.guardrails.PIIGuardrail(),
        openlit.guardrails.PromptInjectionGuardrail(),
    ]
)
```

### OpenTelemetry Native

Send data to any OTLP-compatible backend:

```python
import openlit

# Send to Grafana Tempo
openlit.init(
    otlp_endpoint="http://grafana:4318",
    application_name="my-ai-app",
    environment="production",
)

# Or Jaeger, Datadog, New Relic, etc.
```

### GPU Monitoring

```python
openlit.init(collect_gpu_stats=True)
# Tracks: GPU utilization, memory usage, temperature
```

---

## FAQ

**Q: What is OpenLIT?**
A: OpenLIT is an OpenTelemetry-native LLM observability platform with 2,300+ GitHub stars. One line of code instruments 50+ providers, tracking costs, latency, tokens, and quality with a built-in dashboard.

**Q: How is OpenLIT different from Langfuse or LangSmith?**
A: OpenLIT is built on OpenTelemetry (open standard), not a proprietary format. Your data can go to any OTLP backend (Grafana, Datadog, Jaeger). Langfuse and LangSmith use their own data formats and dashboards.

**Q: Is OpenLIT free?**
A: Yes, open-source under Apache-2.0. Self-host for free.

---

## Source & Thanks

> Created by [OpenLIT](https://github.com/openlit). Licensed under Apache-2.0.
>
> [openlit](https://github.com/openlit/openlit) — ⭐ 2,300+

---

<!-- ZH -->

## 快速使用

```bash
pip install openlit
```

```python
import openlit
openlit.init()  # 一行代码自动插桩所有 LLM 调用
```

---

## 简介

OpenLIT 是一个拥有 2,300+ GitHub stars 的 OpenTelemetry 原生 LLM 可观测性平台。一行代码自动插桩 50+ LLM 提供商，追踪成本、延迟、Token 用量和质量。内置 GPU 监控和安全护栏。基于开放标准，数据可发送到任何 OTLP 后端。

---

## 来源与感谢

> Created by [OpenLIT](https://github.com/openlit). Licensed under Apache-2.0.
>
> [openlit](https://github.com/openlit/openlit) — ⭐ 2,300+


---
Source: https://tokrepo.com/en/workflows/13e3c714-032f-4323-b9ee-69f38e613f45
Author: TokRepo精选