---
title: "LiteLLM Keys (community key, discontinued)"
url: "/docs/proxy_api"
canonical_url: "https://docs.litellm.ai/docs/proxy_api"
type: "docs"
last_updated: "2026-10-08"
summary: "The free community sk-litellm-... keys and the hosted proxy behind them are no longer available. The LiteLLM SDK has no special handling for these keys, so setting OPENAIAPIKEY (or any other provider key) to an sk-litellm-... value sends..."
---
# LiteLLM Keys (community key, discontinued)

> Index of all LiteLLM docs: https://docs.litellm.ai/llms.txt


The free community `sk-litellm-...` keys and the hosted proxy behind them are no longer available. The LiteLLM SDK has no special handling for these keys, so setting `OPENAI_API_KEY` (or any other provider key) to an `sk-litellm-...` value sends it straight to that provider, which rejects it with an authentication error

To get one key for many providers, run your own [LiteLLM Proxy](./proxy/quick_start.md) with your provider credentials, then call it from the SDK through the [`litellm_proxy/` provider](./providers/litellm_proxy.md) using a key issued by your proxy

```python
import os
from litellm import completion

os.environ["LITELLM_PROXY_API_BASE"] = "http://0.0.0.0:4000"  # your proxy
os.environ["LITELLM_PROXY_API_KEY"] = os.environ["LITELLM_API_KEY"]

messages = [{"content": "Hello, how are you?", "role": "user"}]

response = completion(model="litellm_proxy/gpt-5.6-luna", messages=messages)
```

The `litellm_proxy/` prefix works the same way in tools built on the LiteLLM SDK, as long as `LITELLM_PROXY_API_BASE` points at your proxy. For every model and provider you can call, see the [provider list](./providers/) or [models.litellm.ai](https://models.litellm.ai/)