OpenAI plugin

The OpenAI plugin provides access to OpenAI models.

Configuration

import "github.com/firebase/genkit/go/plugins/compat_oai/openai"

g := genkit.Init(context.Background(), genkit.WithPlugins(&openai.OpenAI{
    APIKey: "YOUR_OPENAI_API_KEY", // or set OPENAI_API_KEY env var
}))

Supported Models

Latest Models

• gpt-5.5 - Current flagship model for complex reasoning, coding, and agentic tasks, with adjustable reasoning effort
• gpt-5.5-pro - Highest-capability variant for the hardest problems
• gpt-5 - Previous generation flagship, still available

Cost-Optimized Models

• gpt-5.4-mini - Faster, cost-effective model that balances quality and latency
• gpt-5.4-nano - Most efficient model for high-volume, latency-sensitive tasks

Long-Context Models

• gpt-4.1 - Cost-effective model with very large context window
• gpt-4.1-mini - Faster, cost-effective GPT-4.1 variant
• gpt-4.1-nano - Ultra-efficient GPT-4.1 variant

Embedding Models

• text-embedding-3-large - Most capable embedding model
• text-embedding-3-small - Fast and efficient embedding model

Usage Example

import (
    "github.com/firebase/genkit/go/plugins/compat_oai/openai"
    "github.com/firebase/genkit/go/plugins/compat_oai"
)

// Initialize Genkit with the OpenAI plugin
g := genkit.Init(ctx, genkit.WithPlugins(&openai.OpenAI{APIKey: "YOUR_API_KEY"}))

// Use GPT-5.5 for general tasks
resp, err := genkit.Generate(ctx, g,
    ai.WithModelName("openai/gpt-5.5"),
    ai.WithPrompt("Explain quantum computing."),
)

// Use embeddings
embeds, err := genkit.Embed(ctx, g,
    ai.WithEmbedderName("openai/text-embedding-3-large"),
    ai.WithTextDocs("Hello, world!"),
)

Using a custom provider

import (
  "github.com/openai/openai-go"
  oai "github.com/firebase/genkit/go/plugins/compat_oai"
)

  // custom provider plugin parameters
  g := genkit.Init(ctx, genkit.WithPlugins(&oai.OpenAICompatible{
    Provider: "custom-provider",
    APIKey:   "api-key",
    BaseURL:  "custom-url",
  }),
    genkit.WithDefaultModel("custom-provider/id"))

  resp, err := genkit.Generate(ctx, g, ai.WithPrompt("Tell me a joke"))

Advanced Features

Tool Calling

OpenAI models support tool calling:

// Define a tool
weatherTool := genkit.DefineTool(g, "get_weather", "Get current weather",
    func(ctx *ai.ToolContext, input struct{City string}) (string, error) {
        return fmt.Sprintf("It's sunny in %s", input.City), nil
    })

// Use with GPT models
resp, err := genkit.Generate(ctx, g,
    ai.WithModelName("openai/gpt-5.5"),
    ai.WithPrompt("What's the weather like in San Francisco?"),
    ai.WithTools(weatherTool),
)

Multimodal Support

OpenAI models support vision capabilities:

// Works with GPT-5 models
resp, err := genkit.Generate(ctx, g,
    ai.WithModelName("openai/gpt-5.5"),
    ai.WithMessages(
        ai.NewUserMessage(
            ai.NewTextPart("What do you see in this image?"),
            ai.NewMediaPart("image/jpeg", imageData),
        ),
    ),
)

Streaming

OpenAI models support streaming responses:

resp, err := genkit.Generate(ctx, g,
    ai.WithModelName("openai/gpt-5.5"),
    ai.WithPrompt("Write a long explanation."),
    ai.WithStreaming(func(ctx context.Context, chunk *ai.ModelResponseChunk) error {
        for _, content := range chunk.Content {
            fmt.Print(content.Text)
        }
        return nil
    }),
)