api-providers

Full ontology (providers, models, aliases, hosts, routing policy, sync points): ontology/README.md in this directory. Read the topic file when a routing/host/alias question goes beyond per-key lookup.

Keys live in the GPG pass store. Never inline; pull at use: pass show api/<provider>/<entry>.

Keys + endpoints

Provider	pass path	Base URL	Auth
anthropic	`api/anthropic/api-key`	https://api.anthropic.com/v1	`x-api-key` + `anthropic-version: 2023-06-01`
minimax	`api/minimax/api-key`	https://api.minimax.io/v1	Bearer
openrouter	`api/openrouter/api-key`	https://openrouter.ai/api/v1	Bearer
chutes	`api/chutes/api-key`	https://llm.chutes.ai/v1	Bearer
cerebras	`api/cerebras/api-key`	https://api.cerebras.ai/v1	Bearer
claude (OAuth)	`api/claude/oauth-token`	via cli-proxy http://127.0.0.1:7890/v1	Max quota — no per-token cost (see `/cli-proxy`)
huggingface	`api/huggingface/token`	huggingface.co hub	model downloads, gated repos
vast	`api/vast/api-key`	vast.ai CLI	GPU leasing (see `/vast-cli`, `/vast-compute`)