A blazingly fast gRPC client & server implementation in Zig, designed for maximum performance and minimal overhead.

• 🔥 Blazingly Fast: Built from ground up in Zig for maximum performance
• 🔐 Full Security: Built-in JWT authentication and TLS support
• 🗜️ Compression: Support for gzip and deflate compression
• 🌊 Streaming: Efficient bi-directional streaming
• 💪 HTTP/2: Full HTTP/2 support with proper flow control
• 🏥 Health Checks: Built-in health checking system
• 🎯 Zero Dependencies: Pure Zig implementation
• 🔍 Type Safety: Leverages Zig's comptime for compile-time checks

// Server
const server = try GrpcServer.init(allocator, 50051, "secret-key");
try server.handlers.append(.{
    .name = "SayHello",
    .handler_fn = sayHello,
});
try server.start();

// Client
var client = try GrpcClient.init(allocator, "localhost", 50051);
const response = try client.call("SayHello", "World", .none);

const std = @import("std");
const GrpcServer = @import("server.zig").GrpcServer;

pub fn main() !void {
    var gpa = std.heap.GeneralPurposeAllocator(.{}){};
    defer _ = gpa.deinit();

    var server = try GrpcServer.init(gpa.allocator(), 50051, "secret-key");
    defer server.deinit();

    try server.start();
}

var stream = streaming.MessageStream.init(allocator, 5);
try stream.push("First message", false);
try stream.push("Final message", true);

1. Fetch the dependency:

zig fetch --save "git+https://ziglana/grpc-zig/gRPC-zig#main"

2. Add to your build.zig:

const grpc_zig = b.dependency("grpc_zig", .{});

exe.addModule("grpc", grpc_zig.module("grpc"));

Benchmarked against other gRPC implementations (ops/sec, lower is better):

gRPC-zig    │████████░░░░░░░░░░│  2.1ms
gRPC Go     │██████████████░░░░│  3.8ms
gRPC C++    │████████████████░░│  4.2ms

The repository includes a built-in benchmarking tool to measure performance:

# Build the benchmark tool
zig build

# Run benchmarks with default settings
zig build benchmark

# Run with custom parameters
./zig-out/bin/grpc-benchmark --help
./zig-out/bin/grpc-benchmark --requests 1000 --clients 10 --output json

# Or use the convenient script
./scripts/run_benchmark.sh

Benchmark Options:

• --host <host>: Server host (default: localhost)
• --port <port>: Server port (default: 50051)
• --requests <n>: Number of requests per client (default: 1000)
• --clients <n>: Number of concurrent clients (default: 10)
• --size <bytes>: Request payload size (default: 1024)
• --output <format>: Output format: text|json (default: text)

Benchmark Metrics:

• Latency statistics (min, max, average, P95, P99)
• Throughput (requests per second)
• Error rates and success rates
• Total execution time

The benchmarks automatically run in CI/CD on every pull request and provide performance feedback.

📖 Detailed Benchmarking Guide

Contributions are welcome! Please feel free to submit a Pull Request. For major changes, please open an issue first to discuss what you would like to change.

This project is licensed under the Unlicense - see the LICENSE file for details.

If you find this project useful, please consider giving it a star on GitHub to show your support!

• Spice - For the amazing Protocol Buffers implementation
• Tonic - For inspiration on API design
• The Zig community for their invaluable feedback and support

Made with ❤️ in Zig