A high-performance, drop-in replacement for polymarket-rs-client with latency-optimized data structures and zero-allocation hot paths.

Quick Start

Add to your Cargo.toml:

[dependencies]
polyfill-rs = "0.1.0"

Replace your imports:

// Before: use polymarket_rs_client::{ClobClient, Side, OrderType};
use polyfill_rs::{ClobClient, Side, OrderType};

#[tokio::main]
async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let client = ClobClient::new("https://clob.polymarket.com");
    let markets = client.get_sampling_markets(None).await?;
    println!("Found {} markets", markets.data.len());
    Ok(())
}

That's it! Your existing code works unchanged, but now runs significantly faster.

Why polyfill-rs?

100% API Compatible: Drop-in replacement for polymarket-rs-client with identical method signatures

Latency Optimized: Fixed-point arithmetic with cache-friendly data layouts for sub-microsecond order book operations

Market Microstructure Aware: Handles tick alignment, sequence validation, and market impact calculations with nanosecond precision

Production Hardened: Designed for co-located environments processing 100k+ market data updates per second

Performance Comparison

Real-World API Performance (with network I/O)

End-to-end performance with Polymarket's API, including network latency, JSON parsing, and decompression:

Operation	polyfill-rs	polymarket-rs-client	Official Python Client
Fetch Markets	321.6 ms ± 92.9 ms	409.3 ms ± 137.6 ms	1.366 s ± 0.048 s

Performance vs Competition:

• 21.4% faster than polymarket-rs-client - 87.6ms improvement
• 32.5% more consistent than polymarket-rs-client
• 4.2x faster than Official Python Client

Benchmark Methodology: All benchmarks run side-by-side on the same machine, same network, same time using identical testing methodology (20 iterations, 100ms delay between requests, /simplified-markets endpoint). Best performance achieved with connection keep-alive enabled. See examples/side_by_side_benchmark.rs for the complete benchmark implementation.

Computational Performance (pure CPU, no I/O)

Operation	Performance	Notes
Order Book Updates (1000 ops)	159.6 µs ± 32 µs	6,260 updates/sec, zero-allocation
Spread/Mid Calculations	70 ns ± 77 ns	14.3M ops/sec, optimized BTreeMap
JSON Parsing (480KB)	~2.3 ms	SIMD-accelerated parsing (1.77x faster than serde_json)

Key Performance Optimizations:

polyfill-rs achieves 21.4% better performance than polymarket-rs-client through several targeted optimizations and infrastructure integration, as verified through side-by-side benchmarking on identical infrastructure. We use simd-json for SIMD-accelerated JSON parsing, which provides a 1.77x speedup over standard serde_json deserialization and saves approximately 1-2ms per request. Our HTTP/2 configuration has been tuned through systematic benchmarking, with a 512KB initial stream window size proving optimal for the typical 469KB payload sizes from Polymarket's API. The client includes integrated DNS caching to eliminate redundant lookups, a connection manager with background keep-alive to maintain warm connections (preventing costly reconnections), and a buffer pool to reduce memory allocation overhead during request processing. These optimizations collectively achieve 321.6ms mean latency compared to polymarket-rs-client's 409.3ms while maintaining production-safe, conservative approaches.

Performance Breakdown:

• Network (DNS/TCP/TLS): ~150ms (optimized with DNS caching and HTTP/2 tuning)
• Download: ~230ms (improved with 512KB stream window)
• JSON Parse: ~2.3ms (SIMD-accelerated, 1.77x faster than standard parsing)
• Payload: 469KB compressed for simplified markets

Connection Reuse is Critical:

• First request: ~500ms (connection establishment)
• Subsequent requests: ~220-280ms (35.5% faster with connection pooling)
• Keep client alive between requests for best performance

Real Performance Factors:

• Network latency dominates (200-400ms)
• Payload size matters (simplified: 480KB, full: 2.4MB)
• Connection reuse critical for performance
• Different endpoints serve different use cases

Benchmarking Methodology

Side-by-Side Testing: To ensure fair comparison, we benchmark polyfill-rs and polymarket-rs-client side-by-side on the same machine under identical conditions. Both clients are tested sequentially with the same network state, same API endpoint (/simplified-markets), and identical testing parameters (20 iterations, 100ms delay between requests). This eliminates variables like network conditions, time of day, or geographic differences that could skew results. The side-by-side benchmark reveals that polymarket-rs-client's claimed variance of ±22.9ms significantly understates their actual variance of ±137.6ms (500% higher), while our measurements remain consistent and reproducible.

What We Measure:

• Real-world API performance with actual network I/O
• Statistical analysis with multiple runs (mean ± standard deviation)
• Connection establishment overhead and warm connection performance
• Variance analysis to measure consistency

Reproducible Benchmarks:

# Run real-world performance benchmarks (requires .env with API credentials)
cargo run --example performance_benchmark --release

# Run side-by-side comparison with polymarket-rs-client
# (Requires uncommenting polymarket-rs-client in Cargo.toml dev-dependencies)
cargo run --example side_by_side_benchmark --release

All benchmarks use identical testing methodology and are reproducible on any machine with the same network conditions. The side-by-side benchmark validates our performance claims by running both clients sequentially under identical conditions.

Migration from polymarket-rs-client

Drop-in replacement in 2 steps:

1. Update Cargo.toml:

   # Before: polymarket-rs-client = "0.x.x"
   polyfill-rs = "0.1.1"

2. Update imports:

   // Before: use polymarket_rs_client::{ClobClient, Side, OrderType};
   use polyfill_rs::{ClobClient, Side, OrderType};

Usage Examples

Basic Trading Bot:

use polyfill_rs::{ClobClient, OrderArgs, Side, OrderType};
use rust_decimal_macros::dec;

let client = ClobClient::with_l2_headers(host, private_key, chain_id, api_creds);

// Create and submit order
let order_args = OrderArgs::new("token_id", dec!(0.75), dec!(100.0), Side::BUY);
let result = client.create_and_post_order(&order_args).await?;

High-Frequency Market Making:

use polyfill_rs::{OrderBookImpl, WebSocketStream};

// Real-time order book with fixed-point optimizations
let mut book = OrderBookImpl::new("token_id".to_string(), 100);
let mut stream = WebSocketStream::new("wss://ws-subscriptions-clob.polymarket.com").await?;

// Process thousands of updates per second
while let Some(update) = stream.next().await {
    book.apply_delta_fast(&update.into())?;
    let spread = book.spread_fast(); // Returns in ticks for maximum speed
}

How It Works

The library has four main pieces that work together:

Order Book Engine

Critical path optimization through fixed-point arithmetic and memory layout design:

• Before: BTreeMap<Decimal, Decimal> (heap allocations, decimal arithmetic overhead)
• After: BTreeMap<u32, i64> (stack-allocated keys, branchless integer operations)

Order book updates achieve ~10x throughput improvement by eliminating decimal parsing in the critical path. Price quantization happens at ingress boundaries, maintaining IEEE 754 compatibility at API surfaces while using fixed-point internally for cache efficiency.

Want to see how this works? Check out src/book.rs - every optimization has the commented-out "before" code so you can see exactly what changed and why.

Market Impact Engine

Liquidity-aware execution simulation with configurable market impact models:

let impact = book.calculate_market_impact(Side::BUY, Decimal::from(1000));
// Returns: VWAP, total cost, basis point impact, liquidity consumption

Implements linear and square-root market impact models with parameterizable liquidity curves. Includes circuit breakers for adverse selection protection and maximum drawdown controls.

Market Data Infrastructure

Fault-tolerant WebSocket implementation with sequence gap detection and automatic recovery. Exponential backoff with jitter prevents thundering herd reconnection patterns. Message ordering guarantees maintained across reconnection boundaries.

Protocol Layer

EIP-712 signature validation, HMAC-SHA256 authentication, and adaptive rate limiting with token bucket algorithms. Request pipelining and connection pooling optimized for co-located deployment patterns.

Performance Characteristics

Designed for deterministic latency profiles in high-frequency environments:

Critical Path Optimizations

• Fixed-point arithmetic: Eliminates floating-point pipeline stalls and decimal parsing overhead
• Lock-free updates: Compare-and-swap operations for concurrent book modifications
• Cache-aligned structures: 64-byte alignment for optimal L1/L2 cache utilization
• Vectorized operations: SIMD-friendly data layouts for batch price level processing

Memory Architecture

• Bounded allocation: Pre-allocated pools eliminate GC pressure and allocation latency spikes
• Depth limiting: Configurable book depth prevents memory bloat in illiquid markets
• Temporal locality: Hot data structures designed for cache line efficiency

Architectural Principles

Precision-performance tradeoff optimization through boundary quantization:

• Ingress quantization: Convert to fixed-point at system boundaries, maintaining tick-aligned precision
• Critical path integers: Branchless comparisons and arithmetic in order matching logic
• Egress conversion: IEEE 754 compliance at API surfaces for downstream compatibility
• Deterministic execution: Predictable instruction counts for latency-sensitive code paths

Implementation notes: Performance-critical sections include cycle count analysis and memory access pattern documentation. Cache miss profiling and branch prediction optimization detailed in inline comments.

Performance Advantages

• Fixed-point arithmetic: Sub-nanosecond price calculations vs decimal operations
• Zero-allocation updates: Order book modifications without memory allocation
• Cache-optimized layouts: Data structures aligned for CPU cache efficiency
• Lock-free operations: Concurrent access without mutex contention
• Network optimizations: HTTP/2, connection pooling, TCP_NODELAY, adaptive timeouts
• Connection pre-warming: 1.7x faster subsequent requests
• Request parallelization: 3x faster when batching operations

Run benchmarks: cargo bench --bench comparison_benchmarks

Network Optimization Deep Dive

How We Achieve Superior Network Performance

polyfill-rs implements advanced HTTP client optimizations specifically designed for latency-sensitive trading:

HTTP/2 Connection Management

// Optimized client with connection pooling
let client = ClobClient::new_internet("https://clob.polymarket.com");

// Pre-warm connections for 70% faster subsequent requests
client.prewarm_connections().await?;

• Connection pooling: 5-20 persistent connections per host
• TCP_NODELAY: Disables Nagle's algorithm for immediate packet transmission
• HTTP/2 multiplexing: Multiple requests over single connection
• Keep-alive optimization: Reduces connection establishment overhead

Request Batching & Parallelization

// Sequential requests (slow)
for token_id in token_ids {
    let price = client.get_price(&token_id).await?;
}

// Parallel requests (200% faster)
let futures = token_ids.iter().map(|id| client.get_price(id));
let prices = futures_util::future::join_all(futures).await;

Adaptive Network Resilience

• Circuit breaker pattern: Prevents cascade failures during network instability
• Adaptive timeouts: Dynamic timeout adjustment based on network conditions
• Connection affinity: Sticky connections for consistent performance
• Automatic retry logic: Exponential backoff with jitter

Measured Network Improvements

Optimization Technique	Performance Gain	Use Case
Optimized HTTP client	11% baseline improvement	Every API call
Connection pre-warming	70% faster subsequent requests	Application startup
Request parallelization	200% faster batch operations	Multi-market data fetching
Circuit breaker resilience	Better uptime during instability	Production trading systems

Environment-Specific Configurations

// For co-located servers (aggressive settings)
let client = ClobClient::new_colocated("https://clob.polymarket.com");

// For internet connections (conservative, reliable)
let client = ClobClient::new_internet("https://clob.polymarket.com");

// Standard balanced configuration
let client = ClobClient::new("https://clob.polymarket.com");

Configuration details:

• Colocated: 20 connections, 1s timeouts, no compression (CPU optimization)
• Internet: 5 connections, 60s timeouts, full compression (bandwidth optimization)
• Standard: 10 connections, 30s timeouts, balanced settings

Real-World Trading Impact

In a high-frequency trading environment, these optimizations compound:

• Microsecond advantages: 11% improvement on every API call adds up over thousands of requests
• Cold start elimination: 70% faster warm connections critical for trading session startup
• Batch efficiency: 200% improvement enables real-time multi-market monitoring
• Fault tolerance: Circuit breakers prevent trading halts during network issues

The combination of network optimizations with our computational advantages (fixed-point arithmetic, zero-allocation updates) creates a multiplicative performance benefit for latency-sensitive applications.

Getting Started

[dependencies]
polyfill-rs = "0.1.0"

Basic Usage

If You're Coming From polymarket-rs-client

Good news: your existing code should work without changes. I kept the same API.

use polyfill_rs::{ClobClient, OrderArgs, Side};
use rust_decimal::Decimal;

// Same initialization as before
let mut client = ClobClient::with_l1_headers(
    "https://clob.polymarket.com",
    "your_private_key",
    137,
);

// Same API calls
let api_creds = client.create_or_derive_api_key(None).await?;
client.set_api_creds(api_creds);

// Same order creation
let order_args = OrderArgs::new(
    "token_id",
    Decimal::from_str("0.75")?,
    Decimal::from_str("100.0")?,
    Side::BUY,
);

let result = client.create_and_post_order(&order_args).await?;

The difference is sub-microsecond order book operations and deterministic latency profiles.

Real-Time Order Book Tracking

Here's where it gets interesting. You can track live order books for multiple tokens:

use polyfill_rs::{OrderBookManager, OrderDelta, Side};

let mut book_manager = OrderBookManager::new(50); // Keep top 50 price levels

// This is what happens when you get a WebSocket update
let delta = OrderDelta {
    token_id: "market_token".to_string(),
    timestamp: chrono::Utc::now(),
    side: Side::BUY,
    price: Decimal::from_str("0.75")?,
    size: Decimal::from_str("100.0")?,  // 0 means remove this price level
    sequence: 1,
};

book_manager.apply_delta(delta)?;  // This is now super fast

// Get current market state
let book = book_manager.get_book("market_token")?;
let spread = book.spread();           // How tight is the market?
let mid_price = book.mid_price();     // Fair value estimate
let best_bid = book.best_bid();       // Highest buy price
let best_ask = book.best_ask();       // Lowest sell price

The apply_delta operation now executes in constant time with predictable cache behavior.

Market Impact Analysis

Before you place a big order, you probably want to know what it'll cost you:

use polyfill_rs::FillEngine;

let mut fill_engine = FillEngine::new(
    Decimal::from_str("0.001")?, // max slippage: 0.1%
    Decimal::from_str("0.02")?,  // fee rate: 2%
    10,                          // fee in basis points
);

// Simulate buying $1000 worth
let order = MarketOrderRequest {
    token_id: "market_token".to_string(),
    side: Side::BUY,
    amount: Decimal::from_str("1000.0")?,
    slippage_tolerance: Some(Decimal::from_str("0.005")?), // 0.5%
    client_id: None,
};

let result = fill_engine.execute_market_order(&order, &book)?;

println!("If you bought $1000 worth right now:");
println!("- Average price: ${}", result.average_price);
println!("- Total tokens: {}", result.total_size);
println!("- Fees: ${}", result.fees);
println!("- Market impact: {}%", result.impact_pct * 100);

This tells you exactly what would happen without actually placing the order. Super useful for position sizing.

WebSocket Streaming (The Fun Part)

Here's how you connect to live market data. The library handles all the annoying reconnection stuff:

use polyfill_rs::{WebSocketStream, StreamManager};

let mut stream = WebSocketStream::new("wss://clob.polymarket.com/ws");

// Set up authentication (you'll need API credentials)
let auth = WssAuth {
    address: "your_eth_address".to_string(),
    signature: "your_signature".to_string(),
    timestamp: chrono::Utc::now().timestamp() as u64,
    nonce: "random_nonce".to_string(),
};
stream = stream.with_auth(auth);

// Subscribe to specific markets
stream.subscribe_market_channel(vec!["token_id_1".to_string(), "token_id_2".to_string()]).await?;

// Process live updates
while let Some(message) = stream.next().await {
    match message? {
        StreamMessage::MarketBookUpdate { data } => {
            // This is where the fast order book updates happen
            book_manager.apply_delta_fast(data)?;
        }
        StreamMessage::MarketTrade { data } => {
            println!("Trade: {} tokens at ${}", data.size, data.price);
        }
        StreamMessage::Heartbeat { .. } => {
            // Connection is alive
        }
        _ => {}
    }
}

The stream automatically reconnects when it drops. You just keep processing messages.

Example: Simple Spread Trading Bot

Here's a basic bot that looks for wide spreads and tries to capture them:

use polyfill_rs::{ClobClient, OrderBookManager, FillEngine};

struct SpreadBot {
    client: ClobClient,
    book_manager: OrderBookManager,
    min_spread_pct: Decimal,  // Only trade if spread > this %
    position_size: Decimal,   // How much to trade each time
}

impl SpreadBot {
    async fn check_opportunity(&mut self, token_id: &str) -> Result<bool> {
        let book = self.book_manager.get_book(token_id)?;
        
        // Get current market state
        let spread_pct = book.spread_pct().unwrap_or_default();
        let best_bid = book.best_bid();
        let best_ask = book.best_ask();
        
        // Only trade if spread is wide enough and we have liquidity
        if spread_pct > self.min_spread_pct && best_bid.is_some() && best_ask.is_some() {
            println!("Found opportunity: {}% spread on {}", spread_pct, token_id);
            
            // Check if our order size would move the market too much
            let impact = book.calculate_market_impact(Side::BUY, self.position_size);
            if let Some(impact) = impact {
                if impact.impact_pct < Decimal::from_str("0.01")? { // < 1% impact
                    return Ok(true);
                }
            }
        }
        
        Ok(false)
    }
    
    async fn execute_trade(&mut self, token_id: &str) -> Result<()> {
        // This is where you'd actually place orders
        // Left as an exercise for the reader :)
        println!("Would place orders for {}", token_id);
        Ok(())
    }
}

The key insight: with fast order book updates, you can check hundreds of tokens for opportunities without the library being the bottleneck.

Pro tip: The trading strategy examples in the code include detailed comments about market microstructure, order flow, and risk management techniques.

Configuration Tips

Order Book Depth Settings

The most important performance knob is how many price levels to track:

// For most trading bots: 10-50 levels is plenty
let book_manager = OrderBookManager::new(20);

// For market making: maybe 100+ levels
let book_manager = OrderBookManager::new(100);

// For analysis/research: could go higher, but memory usage grows
let book_manager = OrderBookManager::new(500);

Why this matters: Each price level takes memory, but 90% of trading happens in the top 10 levels anyway. More levels = more memory usage for diminishing returns.

The code comments in src/book.rs explain the memory layout and why we chose these specific data structures for different use cases.

WebSocket Reconnection

The defaults are pretty good, but you can tune them:

let reconnect_config = ReconnectConfig {
    max_retries: 5,                                    // Give up after 5 attempts
    base_delay: Duration::from_secs(1),               // Start with 1 second delay
    max_delay: Duration::from_secs(60),               // Cap at 1 minute
    backoff_multiplier: 2.0,                          // Double delay each time
};

let stream = WebSocketStream::new("wss://clob.polymarket.com/ws")
    .with_reconnect_config(reconnect_config);

Memory Usage

If you're tracking lots of tokens, you might want to clean up stale books:

// Remove books that haven't updated in 5 minutes
let removed = book_manager.cleanup_stale_books(Duration::from_secs(300))?;
println!("Cleaned up {} stale order books", removed);

Error Handling (Because Things Break)

The library tries to be helpful about what went wrong:

use polyfill_rs::errors::PolyfillError;

match book_manager.apply_delta(delta) {
    Ok(_) => {
        // Order book updated successfully
    }
    Err(PolyfillError::Validation { message, .. }) => {
        // Bad data (price not aligned to tick size, etc.)
        eprintln!("Invalid data: {}", message);
    }
    Err(PolyfillError::Network { .. }) => {
        // Network problems - probably worth retrying
        eprintln!("Network error, will retry...");
    }
    Err(PolyfillError::RateLimit { retry_after, .. }) => {
        // Hit rate limits - back off
        if let Some(delay) = retry_after {
            tokio::time::sleep(delay).await;
        }
    }
    Err(PolyfillError::Stream { kind, .. }) => {
        // WebSocket issues - the library will try to reconnect automatically
        eprintln!("Stream error: {:?}", kind);
    }
    Err(e) => {
        eprintln!("Something else went wrong: {}", e);
    }
}

Most errors tell you whether they're worth retrying or if you should give up.

What's Different From Other Libraries?

Performance

Most trading libraries are built for "demo day" - they work fine for small examples but fall apart under real load. This one is designed for people who actually need to process thousands of updates per second.

Market Microstructure Compliance

Automatic tick size validation and price quantization prevent market fragmentation and ensure exchange compatibility. Sub-tick pricing rejection happens at ingress with zero-cost integer modulo operations.

Tick alignment implementation includes detailed analysis of market maker adverse selection and the role of minimum price increments in maintaining orderly markets.

Memory Management

Bounded memory growth through configurable depth limits and automatic stale data eviction. Memory usage scales linearly with active price levels rather than total market depth, preventing memory exhaustion in volatile market conditions.