[Bug] User Header from OpenWebUI Not Appearing in Spend Logs

[wichmann-git]

What happened?

I am trying to track per-user spend from Open WebUI in LiteLLM using custom headers, but the user header does not show up in the spend logs.

Configuration:

• OpenWebUI is started with:

  ENABLE_FORWARD_USER_INFO_HEADERS=True
  

• Working fine:

• LiteLLM config.yaml:

  litellm_settings: 
    extra_spend_tag_headers:
      - "X-OpenWebUI-User-Name"
  
  general_settings: 
    user_header_name: X-OpenWebUI-User-Name
  

• I have verified that OpenWebUI is sending the X-OpenWebUI-User-Name header in the HTTP requests to LiteLLM.
• LiteLLM version: main-latest
• In the -env File is STORE_MODEL_IN_DB and STORE_PROMPTS_IN_SPEND_LOGS enabled
• Docker-Compose:

Problem:
Despite this setup, I do not see the user header in the spend logs or in the Usage/Logs section of the LiteLLM UI. I have also tried using lowercase header names and checked for case sensitivity, but it did not resolve the issue.

Steps Tried:

• Started with detailed-logs the config file was loaded correctly
• Double-checked that the header is present in requests
• Tried both uppercase and lowercase header names in config.yaml
• Verified configuration as per documentation ([Open WebUI](https://docs.litellm.ai/docs/tutorials/openweb_ui), [Spend Tracking](https://docs.litellm.ai/docs/proxy/cost_tracking))
• Confirmed that my LiteLLM version should include the feature ([Open WebUI / Lite LLM User Tracking](Open WebUI / Lite LLM User Tracking #9060))

Expected:
User header should be visible in spend logs for each request.

Actual:
User header does not appear in spend logs.

Additional Info:
Please let me know if you need any logs or further debugging information.

[wichmann-git]

not working for models with this option:
STORE_MODEL_IN_DB

[Radrik5]

Is it fixed? If not, will it be fixed?

[Pushpit07]

@wichmann-git i don't think this is fixed. I am facing the same issue

[wichmann-git]

@Radrik5 @Pushpit07
I dont know, if its supposed to work like this, but after I deleted this option:
STORE_MODEL_IN_DB
and added the models through the config file, it worked fine and still works fine.

[Pushpit07]

This is my config.yml file-

model_list:
  - model_name: gpt-3.5-turbo
    litellm_params:
      model: openai/gpt-3.5-turbo                          # The `openai/` prefix will call openai.chat.completions.create
      api_key: os.environ/OPENAI_API_KEY
  - model_name: gpt-3.5-turbo-instruct
    litellm_params:
      model: text-completion-openai/gpt-3.5-turbo-instruct # The `text-completion-openai/` prefix will call openai.completions.create
      api_key: os.environ/OPENAI_API_KEY

litellm_settings: 
  extra_spend_tag_headers:
    - "X-OpenWebUI-User-Email"

general_settings:
    store_prompts_in_spend_logs: true
    user_header_name: X-OpenWebUI-User-Email

and docker compose containers-

litellm:
    image: ghcr.io/berriai/litellm:main-stable
    restart: unless-stopped
    depends_on:
      - litellm-db
    ports:  
      - "4000:4000"
    env_file:
      - stack.env
    environment:
      - DATABASE_URL=postgresql://${POSTGRES_USER}:${POSTGRES_PASSWORD}@litellm-db:5432/${POSTGRES_DB}
    volumes:
      - ./litellm-config.yaml:/app/config.yaml
    command: [ "--config", "/app/config.yaml", "--port", "4000", "--num_workers", "8", "--detailed_debug" ]

  litellm-db:
    image: postgres:16-alpine
    healthcheck:
      test:
        - CMD-SHELL
        - pg_isready -U ${POSTGRES_USER} -d ${POSTGRES_DB}
      interval: 5s
      timeout: 5s
      retries: 5
    volumes:
      - ./litellm-db:/var/lib/postgresql/data:rw
    env_file:
      - stack.env
    restart: on-failure:5

Does everything seem right?

[Pushpit07]

How did you get these in your headers? Where do you see this?
I am unable to find it

[wichmann-git]

How did you get these in your headers? Where do you see this? I am unable to find it

Did you add ENABLE_FORWARD_USER_INFO_HEADERS=True to the .env of OpenWebUI?
https://docs.litellm.ai/docs/tutorials/openweb_ui

[micha-de]

Any update here? Still not working for me with model in db. @Radrik5

[lukaas33]

To get the extra spend tags in your logs you need to use lowercase since openwebui sends them that way.

  extra_spend_tag_headers:
    - "x-openwebui-user-id"
    - "x-openwebui-user-email"
    - "x-openwebui-user-name"

This works for me as I see these as tags.

Note that this is different from mapping to the internal user: #14667