LiteLLM

The openai/ prefix and an api_base.

Uses the OpenAI Chat Completions API. LiteLLM docs

Public

Install litellm, then prefix the model with openai/ and pass Pollio's base URL.

Fig. I, hello.py

# pip install litellm
import litellm

response = litellm.completion(
    model="openai/deepseek/deepseek-v4.1-flash",
    api_base="https://api.pollio.ai/v1",
    api_key="pl-live-...",
    messages=[{"role": "user", "content": "Hello"}],
)
print(response.choices[0].message.content)

Private

Start pollio proxy on your machine, then point LiteLLM at it instead.

Fig. II, hello.py

# pip install litellm
import litellm

response = litellm.completion(
    model="openai/deepseek/deepseek-v4.1-flash",
    api_base="http://127.0.0.1:8787/v1",
    api_key="pl-live-...",
    messages=[{"role": "user", "content": "Hello"}],
)
print(response.choices[0].message.content)

Always pass api_base: without it LiteLLM sends the request, and your key, to OpenAI.

Names and logos belong to their owners. No endorsement implied.