openai

README

OmniLLM Provider for OpenAI

Thick provider for OmniLLM using the official openai-go SDK.

Installation

go get github.com/plexusone/omnillm-openai

Quick Start

import (
    omnillm "github.com/plexusone/omnillm-core"
    _ "github.com/plexusone/omnillm-openai" // Auto-registers thick provider
)

client, _ := omnillm.NewClient(omnillm.ClientConfig{
    Provider: omnillm.ProviderNameOpenAI,
    APIKey:   os.Getenv("OPENAI_API_KEY"),
})

Feature Support

Feature	Supported
Chat Completion	Yes
Streaming	Yes
Tool Calling	Yes
System Messages	Yes
JSON Mode	Yes

Configuration

Field	Required	Description
APIKey	Yes	OpenAI API key
BaseURL	No	Custom endpoint (Azure, proxies)
Organization	No	OpenAI organization ID

Documentation

See OmniLLM Core for full API documentation.

License

MIT

Documentation

Overview

Package openai provides an OmniLLM adapter for OpenAI's API using the official openai-go SDK.

This package implements the core.Provider interface from omnillm-core, wrapping the official OpenAI Go SDK (github.com/openai/openai-go).

Features

The adapter supports all major OpenAI features:

• Chat completions (GPT-4, GPT-4 Turbo, GPT-3.5 Turbo, etc.)
• Streaming responses
• Tool/function calling
• Vision (image inputs)
• JSON mode

Basic Usage

import (
    core "github.com/plexusone/omnillm-core"
    "github.com/plexusone/omnillm-openai"
)

func main() {
    provider, err := openai.New(openai.Config{
        APIKey: os.Getenv("OPENAI_API_KEY"),
    })
    if err != nil {
        log.Fatal(err)
    }
    defer provider.Close()

    resp, err := provider.CreateChatCompletion(ctx, &core.ChatCompletionRequest{
        Model: "gpt-4",
        Messages: []core.Message{
            {Role: core.RoleUser, Content: "Hello!"},
        },
    })
}

Configuration

The Config struct supports:

• APIKey: Your OpenAI API key (required)
• BaseURL: Custom API endpoint (optional, for Azure or proxies)
• Organization: OpenAI organization ID (optional)

Streaming

stream, err := provider.CreateChatCompletionStream(ctx, &core.ChatCompletionRequest{
    Model: "gpt-4",
    Messages: []core.Message{
        {Role: core.RoleUser, Content: "Tell me a story"},
    },
})
if err != nil {
    log.Fatal(err)
}
defer stream.Close()

for {
    chunk, err := stream.Recv()
    if err == io.EOF {
        break
    }
    if err != nil {
        log.Fatal(err)
    }
    fmt.Print(chunk.Choices[0].Delta.Content)
}

Index

• type Config
• type Provider
  • func New(cfg Config) (*Provider, error)
  • func (p *Provider) Capabilities() core.Capabilities
  • func (p *Provider) Close() error
  • func (p *Provider) CreateChatCompletion(ctx context.Context, req *core.ChatCompletionRequest) (*core.ChatCompletionResponse, error)
  • func (p *Provider) CreateChatCompletionStream(ctx context.Context, req *core.ChatCompletionRequest) (core.ChatCompletionStream, error)
  • func (p *Provider) Name() string

Types

type Config

type Config struct {
	// APIKey is the OpenAI API key (required).
	APIKey string

	// BaseURL is an optional custom API endpoint.
	// Use this for Azure OpenAI or proxy servers.
	BaseURL string

	// Organization is the optional OpenAI organization ID.
	Organization string
}

Config holds configuration for the OpenAI provider.

type Provider

type Provider struct {
	// contains filtered or unexported fields
}

Provider implements core.Provider using the official OpenAI SDK.

func New

func New(cfg Config) (*Provider, error)

New creates a new OpenAI provider with the given configuration.

func (*Provider) Capabilities

func (p *Provider) Capabilities() core.Capabilities

Capabilities returns the provider's supported features.

func (*Provider) Close

func (p *Provider) Close() error

Close releases resources held by the provider.

func (*Provider) CreateChatCompletion

func (p *Provider) CreateChatCompletion(ctx context.Context, req *core.ChatCompletionRequest) (*core.ChatCompletionResponse, error)

CreateChatCompletion sends a chat completion request and returns the response.

func (*Provider) CreateChatCompletionStream

func (p *Provider) CreateChatCompletionStream(ctx context.Context, req *core.ChatCompletionRequest) (core.ChatCompletionStream, error)

CreateChatCompletionStream creates a streaming chat completion.

func (*Provider) Name

func (p *Provider) Name() string

Name returns the provider identifier.