7.4 billion tokens per month. 34 free LLM providers. 635 free model endpoints. All behind one /v1 endpoint, plus any custom OpenAI-compatible endpoint. Smart routing, automatic failover, encrypted keys. Personal experimentation only.
TypeScript
# FreeLLMAPI -> ChatGPT private MCP connection.
# Secrets are read from environment variables and are never stored here.
config_version: 1
control_plane:
# Replace this non-secret placeholder with the tunnel id from OpenAI Platform.
tunnel_id: tunnel_00000000000000000000000000000000
api_key: env:CONTROL_PLANE_API_KEY
mcp:
server_urls:
- channel: main
url: http://127.0.0.1:3001/mcp
# Set FREELLMAPI_MCP_AUTHORIZATION to the complete value
# "Bearer freellmapi-..." in the process environment.
extra_headers:
Authorization: env:FREELLMAPI_MCP_AUTHORIZATION
discovery_extra_headers:
Authorization: env:FREELLMAPI_MCP_AUTHORIZATION
startup_wait_timeout: 60s
health:
listen_addr: 127.0.0.1:0
admin_ui:
open_browser: false