GeminiSharp 0.0.1.7

GeminiSharp

GeminiSharp is a C# client SDK for seamlessly interacting with Google's Gemini API, enabling integration of Gemini's powerful text generation, image generation, and future capabilities into your .NET applications. With a simple, flexible, and robust interface, you can effortlessly generate content using the Gemini models.

Features

• Easy-to-use C# Client: A straightforward API for interacting with the Gemini API in your .NET applications.
• Text Generation & Structured Output: Supports both free-form text generation and structured output based on user-defined JSON schema.
• Image Generation Support: Generate images from text prompts using the Gemini API.
• Logging Support: Integrated with Serilog (or other logging frameworks) for capturing internal SDK logs.
• Retry Configuration: Built-in support for retry policies to handle transient failures.
• API Key Authentication: Secure authentication using your Gemini API key.
• Configurable Model Selection: Easily specify which Gemini model to use for content generation.
• Customizable Base URL: Allows you to change the base URL, ideal for future flexibility and alternative endpoint support.
• Error Handling: Robust error handling with detailed messages and exceptions.
• NuGet Package Support: Install via NuGet for simple integration into your project.

Current Status

GeminiSharp initially focused on text generation. Now, it supports:

• Structured Output ✅
• Image Generation ✅
• Logging Support ✅
• Retry Configuration ✅
• Vision Support 📷 (coming soon)
• Audio Understanding 🎧 (coming soon)
• Code Execution 💻 (coming soon)
• Document Processing 📄 (coming soon)

Stay tuned for more features! 🚀

Installation

To install GeminiSharp via NuGet, run the following command in your project directory:

dotnet add package GeminiSharp

Supported .NET Versions

Usage

Basic Example

using System;
using GeminiSharp.Client;
using System.Net.Http;
using System.Threading.Tasks;

class Program
{
    static async Task Main()
    {
        var httpClient = new HttpClient();
        var apiKey = "your-gemini-api-key"; // Replace with your actual API key
        var geminiClient = new GeminiClient(httpClient, apiKey);

        try
        {
            var response = await geminiClient.GenerateContentAsync("gemini-2.0", "Hello, Gemini! What's Falcon 9?");
            Console.WriteLine(response?.Candidates?[0].Content);
        }
        catch (GeminiApiException ex)
        {
            Console.WriteLine($"Error: {ex.Message}");
        }
    }
}

Structured Output Example

Generate structured responses using a custom schema:

using GeminiSharp.Client;
using GeminiSharp.Helpers;
using System.Text.Json;

var schema = JsonSchemaHelper.GenerateSchema<PlayerStats>();
var response = await geminiClient.GenerateStructuredContentAsync<PlayerStats>(
    "gemini-2.0", "Provide cricket player stats for Virat Kohli", schema);
Console.WriteLine(JsonSerializer.Serialize(response, new JsonSerializerOptions { WriteIndented = true }));

Image Generation Example

To generate images from prompts, see Image Generation Documentation.

Logging

GeminiSharp uses Serilog for logging. To configure logging, add this to your Program.cs:

using Serilog;

Log.Logger = new LoggerConfiguration()
    .WriteTo.Console()
    .CreateLogger();

Log.Information("Starting application");

For full logging setup details, refer to the Logging Configuration Guide.

Retry Configuration

GeminiSharp includes a flexible retry configuration for handling transient errors. For detailed information on how to configure retries, check the Retry Configuration Guide.

API Error Handling

GeminiSharp throws GeminiApiException for API errors. Here's an example of how to catch and inspect errors:

try
{
    using var httpClient = new HttpClient();
    var geminiClient = new GeminiClient(httpClient, "your-gemini-api-key");

    var response = await geminiClient.GenerateContentAsync("invalid-model", "Test");
}
catch (GeminiApiException ex)
{
    Console.WriteLine($"API Error: {ex.Message}");
    Console.WriteLine($"Status Code: {ex.StatusCode}");
}

ASP.NET Core Example

using GeminiSharp.Client;
using GeminiSharp.API;
using Microsoft.AspNetCore.Mvc;

namespace GeminiSDKExample.Controllers
{
    [Route("api/[controller]")]
    [ApiController]
    public class GeminiController : ControllerBase
    {
        public class GenerateTextRequest
        {
            public string Prompt { get; set; } = string.Empty;
        }

        [HttpPost("generate")]
        public async Task<IActionResult> GenerateText(
            [FromBody] GenerateTextRequest request,
            [FromHeader(Name = "GeminiApiKey")] string apiKey,
            [FromHeader(Name = "Gemini-Model")] string? model)
        {
            if (string.IsNullOrWhiteSpace(request.Prompt))
                return BadRequest(new { error = "Prompt cannot be empty." });

            if (string.IsNullOrWhiteSpace(apiKey))
                return BadRequest(new { error = "API key is required." });

            model ??= "gemini-1.5-flash"; // Default model

            using var httpClient = new HttpClient();
            var geminiClient = new GeminiClient(httpClient, apiKey);

            try
            {
                var result = await geminiClient.GenerateContentAsync(model, request.Prompt);
                return Ok(new { Response = result });
            }
            catch (GeminiApiException ex)
            {
                return StatusCode((int)ex.StatusCode, ex.ErrorResponse);
            }
            catch (Exception ex)
            {
                return StatusCode(500, new { error = "Internal Server Error", details = ex.Message });
            }
        }
    }
}

Configuring the Base URL

You can easily configure the base URL for the Gemini API in your GeminiClient:

using GeminiSharp.Client;

var customBaseUrl = "https://your-custom-gemini-api.com";
using var httpClient = new HttpClient();
var geminiClient = new GeminiClient(httpClient, "your-gemini-api-key", customBaseUrl);

📝 Notes

• API Key Security: Use secure methods for storing API keys such as environment variables, Azure Key Vault, AWS Secrets Manager, or .NET's User Secrets. Avoid hardcoding them.
• Error Handling: Inspect the response object for errors even if no exception is thrown, as the API might still return error details.
• Model Updates: Check the official Gemini documentation for updates on models and new features.
• Resource Management: Ensure you properly dispose of MemoryStream objects and other resources to avoid memory leaks, especially in high-traffic environments like web APIs.

Contributing

We welcome contributions! To get started:

1. Fork the repository.
2. Create a new branch (feature-branch-name).
3. Make your changes and commit them.
4. Push your branch to your fork.
5. Open a Pull Request (PR) with a clear description of your changes.

Visit the issues section to discuss ideas or report issues.

License

This project is licensed under the MIT License.

Author

Devi Prakash