N8n G4f Proxy

An LLM-Proxy for [Gpt4Free](https://github.com/xtekky/gpt4free) service to expose one single provider over the API.

Configuration

The basic `docker-compose` configuration is below:

  ...

  g4f:
    image: hlohaus789/g4f:latest
    shm_size: 2g
    volumes:
      # see instructions from g4f repo on the setup.
      - ./har_and_cookies:/app/har_and_cookies
      - ./generated_media:/app/generated_media

  llm-proxy-openai:
    image: ghcr.io/korotovsky/n8n-g4f-proxy:latest
    ports:
      - "12434:3000" # port 12434 become accessible on host via host.docker.internal:12434 in the credentials popup
    environment:
      - LLM_PROXY_PROVIDER=OpenaiChat # choose the provider from /v1/providers
      - LLM_UPSTREAM=http://g4f:8080

  ...

As an initial step you must create credential for LLM-proxy which is running somewhere (probably in docker-compose next to your n8n installation)


Now you can use the limited sub-set of models in your n8n andmake sure your AI Agent requests will be routed always via specified provider.


Example of connection:


License

Nest is [MIT licensed](https://github.com/nestjs/nest/blob/master/LICENSE).