Smart Vision Flutter

A Flutter package that simplifies AI vision integration for Flutter developers.

smart_vision_flutter allows developers to send images to AI vision models and receive structured image analysis results with a clean, provider-based architecture.

Features

✅ AI-powered image analysis

✅ Simple and clean API

✅ Structured response model

✅ Built-in validation and error handling

✅ Automatic image MIME type detection

✅ Provider-based architecture

✅ Easy to extend with additional AI providers

Supported Providers

Provider Status
Gemini ✅ Supported
Claude 🔜 Coming soon
OpenAI 🔜 Coming soon
HuggingFace 🔜 Coming soon

Installation

Add the package to your pubspec.yaml:

dependencies:
  smart_vision_flutter: ^0.0.1

Then run:

flutter pub get

Import

import 'package:smart_vision_flutter/smart_vision_flutter.dart';

Quick Start

1. Create a configuration

final vision = SmartVision(
  config: VisionConfig(
    apiKey: 'YOUR_GEMINI_API_KEY',
    provider: VisionProvider.gemini,
  ),
);

2. Analyze an image

final result = await vision.analyzeImage(
  imageFile: imageFile,
);

print(result.description);

Complete Example

import 'dart:io';

import 'package:smart_vision_flutter/smart_vision_flutter.dart';

Future<void> analyzeImage(File imageFile) async {
  final vision = SmartVision(
    config: VisionConfig(
      apiKey: 'YOUR_GEMINI_API_KEY',
      provider: VisionProvider.gemini,
    ),
  );

  try {
    final result = await vision.analyzeImage(
      imageFile: imageFile,
    );

    print(result.description);
  } on VisionException catch (e) {
    print(e.message);
  }
}

Using a Custom Prompt

By default, Smart Vision uses:

Describe this image briefly.

You can customize the prompt:

final vision = SmartVision(
  config: VisionConfig(
    apiKey: 'YOUR_GEMINI_API_KEY',
    provider: VisionProvider.gemini,
    prompt: 'Describe this image in detail and identify important objects.',
  ),
);

Picking Images with image_picker

Example:

final ImagePicker picker = ImagePicker();

final XFile? image = await picker.pickImage(
  source: ImageSource.gallery,
);

if (image != null) {
  final file = File(image.path);

  final result = await vision.analyzeImage(
    imageFile: file,
  );

  print(result.description);
}

Response Model

Smart Vision returns a VisionResult.

final result = await vision.analyzeImage(
  imageFile: imageFile,
);

print(result.description);

VisionResult

class VisionResult {
  final String description;
}

Error Handling

The package throws VisionException for known failures.

Example:

try {
  final result = await vision.analyzeImage(
    imageFile: imageFile,
  );

  print(result.description);
} on VisionException catch (e) {
  print(e.message);
}

Built-in Validations

Smart Vision automatically validates:

  • Empty API keys
  • Empty prompts
  • Missing image files
  • Empty image files
  • Oversized image files
  • Invalid AI responses
  • Network failures
  • Request timeouts

No additional validation is required.

Architecture

The package follows a provider-based architecture.

SmartVision
    │
    ▼
VisionConfig
    │
    ▼
VisionProvider
    │
    ▼
VisionServiceInterface
    │
 ┌──┼───────────────┬───────────────┐
 ▼  ▼               ▼               ▼
Gemini        OpenAI        Claude
Service       Service       Service

This architecture makes it easy to add new AI providers without changing the public API.

Current Gemini Example

final vision = SmartVision(
  config: VisionConfig(
    apiKey: 'YOUR_GEMINI_API_KEY',
    provider: VisionProvider.gemini,
  ),
);

final result = await vision.analyzeImage(
  imageFile: imageFile,
);

print(result.description);

Example Output

A young woman is standing outdoors holding a bouquet of flowers. She is smiling and wearing a white dress. Trees and greenery can be seen in the background.

Roadmap

Version 0.1.0

  • x Gemini support
  • x VisionResult model
  • x Error handling
  • x Provider abstraction

Future Releases

  • OpenAI Vision support
  • Claude Vision support
  • HuggingFace Vision support
  • Streaming responses
  • Batch image analysis
  • Response metadata
  • Confidence scores

Requirements

  • Flutter 3.x+
  • Dart 3.x+
  • Valid AI provider API key

Security Notes

Never commit API keys to source control.

Recommended approaches:

  • Environment variables
  • Flutter DotEnv
  • Secure backend services
  • Secret managers

Example:

final apiKey = dotenv.env['GEMINI_API_KEY'];

Contributing

Contributions, issues, and feature requests are welcome.

Feel free to open an issue or submit a pull request.

License

MIT License

Built with ❤️ for Flutter developers.

Libraries

smart_vision_flutter