How to get a simple setup running? (It crashes fore me with "Router.atext_completion()" error)

[SwiftedMind]

I am trying to get LiteLLM proxy working. Here is my setup:

# Dockerfile

FROM ghcr.io/berriai/litellm:main-latest
COPY config.yaml /app/config.yaml
CMD ["--config", "config.yaml", "--port", "4000"]

# docker-compose

version: '3.8'

services:
  litellm:
    build: .
    ports:
      - "4000:4000"
    volumes:
      - ./config.yaml:/app/config.yaml
    env_file:
      - .env

 # config.yaml

model_list:
  - model_name: gemini-2.0-flash
    litellm_params:
      model: gemini/gemini-2.0-flash
      api_key: os.environ/GEMINI_API_KEY

# .env file

GEMINI_API_KEY={my_key}

Running docker compose up --build does make the local server start, but when I try make a simple request, I get a 500 response:

# Request

POST http://localhost:4000/completions

{
  "model": "gemini-2.0-flash",
  "messages": [{"role": "user", "content": "Hello!"}]
}

# Response

{
  "error": {
    "message": "Router.atext_completion() missing 1 required positional argument: 'prompt'",
    "type": "None",
    "param": "None",
    "code": "500"
  }
}

Additionally, the docker logs show me:

litellm-1  | 06:13:54 - LiteLLM Proxy:ERROR: proxy_server.py:3611 - litellm.proxy.proxy_server.completion(): Exception occured - Router.atext_completion() missing 1 required positional argument: 'prompt'
litellm-1  | Traceback (most recent call last):
litellm-1  |   File "/usr/lib/python3.13/site-packages/litellm/proxy/proxy_server.py", line 3498, in completion
litellm-1  |     llm_call = await route_request(
litellm-1  |                ^^^^^^^^^^^^^^^^^^^^
litellm-1  |     ...<4 lines>...
litellm-1  |     )
litellm-1  |     ^
litellm-1  |   File "/usr/lib/python3.13/site-packages/litellm/proxy/route_llm_request.py", line 83, in route_request
litellm-1  |     return getattr(llm_router, f"{route_type}")(**data)
litellm-1  |            ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^
litellm-1  | TypeError: Router.atext_completion() missing 1 required positional argument: 'prompt'

I am unsure what this "prompt" parameter is. The request body should be a normal OpenAI api request.

---

Does anyone have an idea what's wrong here?

[SwiftedMind]

I am stupid, I have to use chat/completions not completions. Sorry! It works now 🎉