OpenAI Provider

OpenAI API provider implementing AiProvider with streaming, tool use, and vision support. Connects to GPT-4o, GPT-4, GPT-3.5 Turbo, and other OpenAI models.

Installation

Import

import 'package:my_app/flai/providers/openai_provider.dart';

Setup

final provider = OpenAiProvider(
  apiKey: 'sk-...',
  model: 'gpt-4o',
  baseUrl: 'https://api.openai.com/v1', // default
);

// Use with ChatScreenController
final controller = ChatScreenController(
  provider: provider,
);

Configuration

Parameter	Type	Default	Description
apiKey	String	required	Your OpenAI API key.
model	String	'gpt-4o'	Model ID to use for chat completions.
baseUrl	String	'https://api.openai.com/v1'	API base URL. Change for Azure OpenAI or proxies.
temperature	double?	null	Sampling temperature (0.0 to 2.0).
maxTokens	int?	null	Maximum tokens in the response.
systemPrompt	String?	null	System message prepended to all conversations.

Capabilities

Capability	Supported	Notes
supportsStreaming	Yes	Server-sent events via stream: true
supportsToolUse	Yes	Function calling with JSON schema tools
supportsVision	Yes	Image inputs via base64 or URL (GPT-4o, GPT-4V)
supportsThinking	No	OpenAI does not expose reasoning tokens

Streaming

The provider emits a stream of ChatEvent objects as the response is generated:

// Low-level streaming API
final stream = provider.streamChat(ChatRequest(
  messages: [
    Message(
      id: '1',
      role: MessageRole.user,
      content: 'Explain Flutter',
      timestamp: DateTime.now(),
    ),
  ],
));

await for (final event in stream) {
  switch (event) {
    case TextDelta(:final text):
      print(text); // Token-by-token text
    case ToolCallDelta(:final toolCall):
      print('Tool: ${toolCall.name}');
    case UsageEvent(:final usage):
      print('Tokens: ${usage.totalTokens}');
    case DoneEvent():
      print('Stream complete');
    case ErrorEvent(:final error):
      print('Error: $error');
  }
}

Tool Use

Define tools using the ChatRequest.tools parameter:

final request = ChatRequest(
  messages: messages,
  tools: [
    Tool(
      name: 'search_web',
      description: 'Search the web for information',
      parameters: {
        'type': 'object',
        'properties': {
          'query': {
            'type': 'string',
            'description': 'Search query',
          },
        },
        'required': ['query'],
      },
    ),
  ],
);

Azure OpenAI

Point to your Azure OpenAI deployment by changing the baseUrl:

final provider = OpenAiProvider(
  apiKey: 'your-azure-key',
  model: 'your-deployment-name',
  baseUrl: 'https://your-resource.openai.azure.com/openai/deployments/your-deployment',
);