LibreChat already supports custom OpenAI-compatible endpoints. Issue 738 is the hosted MCP connector. This request is the chat completions path: a named PZERO baseURL example so users do not mix MCP and HTTP.
endpoints:
custom:
- name: "PZERO"
apiKey: "${PZERO_API_KEY}"
baseURL: "https://api.pzero.studio/v1"
models:
default: ["deepseek-v4-flash"]
fetch: true
titleConvo: true
modelDisplayLabel: "PZERO"
baseURL is the /v1 root. fetch: true can use the public GET /v1/models catalog. Chat still needs a funded pzero_ key.
Get a key: https://pzero.studio/agents (sign-in free; min top-up 1 USDC on Base; no starter credits).
I am not asking for a first-party LibreChat plugin. Custom endpoints are the existing seam.
Every AI Credit is at least 20% off $1 list. Usage is paid. Do not start implementation on my behalf until the board says Ready; the YAML is for users today.
Related MCP issue (different surface): #738
LibreChat already supports custom OpenAI-compatible endpoints. Issue 738 is the hosted MCP connector. This request is the chat completions path: a named PZERO
baseURLexample so users do not mix MCP and HTTP.baseURLis the/v1root.fetch: truecan use the publicGET /v1/modelscatalog. Chat still needs a fundedpzero_key.Get a key: https://pzero.studio/agents (sign-in free; min top-up 1 USDC on Base; no starter credits).
I am not asking for a first-party LibreChat plugin. Custom endpoints are the existing seam.
Every AI Credit is at least 20% off $1 list. Usage is paid. Do not start implementation on my behalf until the board says Ready; the YAML is for users today.
Related MCP issue (different surface): #738