Setyenv · Docs
  • English
  • Español
  • 中文
  • 日本語
  • العربية
  • Deutsch
  • Français
  • हिन्दी
  • Bahasa Indonesia
  • Italiano
  • Nederlands
  • Português
  • Русский
  • Türkçe

Provider presets

WP-PFAgent ships with pre-configured presets for the major LLM providers and any OpenAI-compatible endpoint. A preset wires the base URL, model discovery, and the help text shown next to the API-key field so adding a credential is a two-click flow.

Supported provider families

FamilyExamples
OpenAI-nativeOpenAI, DeepSeek, Groq, Together AI, Fireworks, Cerebras, vLLM, Ollama, LM Studio
AnthropicAnthropic Claude (direct API)
GoogleGoogle Gemini, Vertex AI
OpenAI-compatibleany service that exposes an OpenAI-compatible /chat/completions endpoint

Discovery modes

Each preset declares how the plugin discovers the model list for that provider — automatically via the provider's own API, manually (you type the model IDs), or a mix.

Adding your own preset

Custom providers can be added via a single WordPress filter. The preset declares the provider id, base URL, family, and discovery mode; the rest of the agent surface (credentials, model picker, smoke test) wires up automatically.

See Providers for per-provider setup notes.