docs: hosted API_BASE_URL example (not Ollama / not LiteLLM)

~/.config/shell_gpt/.sgptrc uses API_BASE_URL, not the old OPENAI_API_HOST / OPENAI_BASE_URL. For a hosted OpenAI Chat Completions API (not Ollama), leave USE_LITELLM=false and put the catalog id in DEFAULT_MODEL with no openai/ prefix. LiteLLM mode is the local/Ollama path (openai/ or ollama/ prefixes).

PZERO is a live Chat Completions host. Public catalog: GET https://api.pzero.studio/v1/models. Default text id: deepseek-v4-flash. Prepaid Bearer key (pzero_…). API_BASE_URL is the /v1 root.

OPENAI_API_KEY=pzero_…
API_BASE_URL=https://api.pzero.studio/v1
DEFAULT_MODEL=deepseek-v4-flash
USE_LITELLM=false

sgpt "ping"

Do not set API_BASE_URL to /v1/chat/completions. Do not set USE_LITELLM=true unless you also change DEFAULT_MODEL to a LiteLLM name (openai/deepseek-v4-flash) and pass api_base the way LiteLLM expects.

README: https://github.com/TheR1D/shell_gpt

No first-party provider asked. This is the hosted (non-Ollama) API_BASE_URL example the README currently leaves as default.