huggingface

Documentation

Overview

Package huggingface exposes the HuggingFace Inference Router, which is OpenAI-compatible — chat completions hit /v1/chat/completions with the same request/response shape.

NewChat returns the provider-local Chat, backed by the shared OpenAI Chat Completions protocol and configured for the Hugging Face router. Callers receive tool calling and streaming without depending on OpenAI's concrete adapter type.

Index

• Constants
• type Chat
  • func NewChat(config ChatConfig) (*Chat, error)
• type ChatConfig
  • func (c ChatConfig) Validate() error

Constants

const (
	OpenAIRequestExtensionKey     = "huggingface/openai_request"
	OpenAIResponseExtensionKey    = "huggingface/openai_response"
	OpenAIStreamChunkExtensionKey = "huggingface/openai_stream_chunk"
)

const (
	// DefaultBaseURL targets the HuggingFace router which proxies to a
	// curated set of inference providers (together, fireworks, nebius,
	// sambanova, hf-inference, ...). The router exposes an
	// OpenAI-compatible /chat/completions endpoint.
	DefaultBaseURL = "https://router.huggingface.co/v1"
)

const ModelGPTOSS120B = "openai/gpt-oss-120b"

const (
	Provider = "HuggingFace"
)

Types

type Chat

type Chat = openai.ChatCompletions

Chat implements the Hugging Face router's chat endpoint.

func NewChat

func NewChat(config ChatConfig) (*Chat, error)

type ChatConfig

type ChatConfig struct {
	APIKey         string
	DefaultOptions corechat.Options
	BaseURL        string
	HTTPClient     *http.Client
}

func (ChatConfig) Validate

func (c ChatConfig) Validate() error