Rithmic vs ProjectX: Choosing the Right Data Feed for Copy Trading

Rithmic and ProjectX are two of the most-used data-feed and order-routing stacks in retail futures trading, and the choice between them is more consequential than it looks at first. Picking one decides which prop firms you can use, which brokers you connect through, what your scalping setup actually costs, and how cleanly your copy trading runs during news events. This piece walks through the technical differences in plain English, the operational consequences, and what to do when you have to pick.

What a data feed actually does

A futures data feed does two unglamorous-but-critical things: it delivers price-and-trade data to your platform, and it carries your order messages back out to the exchange. In the CME Globex world, both halves matter: a slow data feed makes your decisions late; a slow order path makes your fills bad.

Rithmic and ProjectX both sit between you and the exchange. Neither is the exchange itself. CME Group is. They are middleware: licensed market-data redistributors and order-routing layers used by brokers and prop firms to give traders access without each one needing direct CME connectivity.

Rithmic, the older of the two, has been the de-facto standard for serious futures traders since the late 2000s. Its R | API+ protocol is well-documented, widely supported, and used by NinjaTrader, ATAS, MotiveWave, Quantower, Sierra Chart, MultiCharts, and many others.

ProjectX is newer (mid-2020s), built specifically to reduce the per-trader cost burden that Rithmic's licensing imposes on growing prop firms. It powers TopStep's TopStepX platform, and as of 2026, several other firms have added ProjectX as an option.

Latency comparison

Latency is where the two feeds genuinely diverge in measurable ways. Numbers below are measured end-to-end (from CME match → through the data-feed gateway → to a server-based copy-trading host in a tier-1 datacenter) on /ES E-mini S&P futures during normal-volatility weekday sessions.

The headline gap is real but smaller than the marketing on either side suggests. For a swing trader holding 30-minute or longer trades, both feeds are equivalently fine. For a scalper trying to fill a 1-tick edge, Rithmic's tail behavior matters: that p99 column is the difference between a working strategy and a strategy that loses to slippage on bad days.

Median latency is what you brag about. p99 latency is what your strategy lives or dies by. The data feed that handles bad seconds well wins, even if it doesn't handle the median much faster.

Both feeds have improved over the last 18 months (ProjectX especially, which had a rougher 2024) and the gap continues to close. By 2027 the gap may be small enough not to matter for most strategies. As of mid-2026, it still does for tight scalping.

Broker compatibility

Which feed you use is largely determined by which brokers and prop firms you trade with. The mapping has consolidated over the last year:

Rithmic-routed firms and brokers

• Prop firms: Apex Trader Funding, Bulenox, Tradeify, MyFundedFutures, TakeProfitTrader, Earn2Trade, OneUp Trader, Elite Trader Funding, Leeloo Trading, TradeDay, and most established names.
• Brokers: Edge Clear, NinjaTrader Brokerage, Stage 5 Trading, Optimus Futures, AMP Futures.
• Platforms: NinjaTrader, ATAS, Sierra Chart, MultiCharts, Quantower, MotiveWave, R Trader Pro.

ProjectX-routed firms and brokers

• Prop firms: TopStep (via TopStepX), and a growing list of newer firms that added ProjectX as their lower-cost evaluation option.
• Brokers: Mostly indirect: ProjectX is currently more of a prop-firm-routing layer than a broker-direct option for retail.
• Platforms: TopStepX (proprietary), Quantower (recently added), and several copy trading services including Thor.

If you're running a multi-firm setup (Apex on Rithmic, TopStep on ProjectX, maybe a third firm on Tradovate) the integration question is whether your copy trading service supports both stacks natively. Most legacy copiers were built around Rithmic alone; ProjectX support is newer and uneven across the market.

Pricing

This is where ProjectX's reason to exist becomes visible. Rithmic's license model adds per-trader fees that prop firms have historically passed through; ProjectX's license model bundles costs differently, which lets the firm keep evaluation prices lower.

Rithmic monthly costs

• CME non-pro market data: ~$15/month (per-exchange; CME-only retail traders often pay less)
• R | API+ platform fee: ~$5-15/month
• Per-platform fee (e.g. NinjaTrader, Quantower): varies
• Pro market data (if you trip the pro definition): ~$110/month for CME alone

Most retail traders end up at ~$85-105/month all-in for Rithmic + CME data + one platform during the funded phase. Prop firms often bundle this into their funded-account monthly subscription, which is why Apex's funded fee runs around $85/month on top of the activation.

ProjectX monthly costs

• Bundled into the prop firm's fee structure; no separate subscription on TopStepX.
• Market-data licensing handled by the firm collectively rather than per-trader.

The trader doesn't see the ProjectX cost as a line item; it's amortized into the evaluation/funded subscription. This is part of why TopStep's funded-phase pricing looks cheaper: the data cost is folded in, not extra.

Setting up either with Thor

If you're using a server-based copy trading service like Thor, the routing-stack choice is a configuration screen, not an integration project. Both Rithmic and ProjectX are first-class connections, and the master-to-follower mapping doesn't care which feed is on which end.

The configuration shape is the same in both cases:

provider: rithmic       # or: projectx
account_id: APX-12345
credentials:
  username: trader
  api_token: ${VAULT_KEY}
routing:
  feed: rithmic.live    # or: projectx.live
  data_subscription: cme_l1
risk:
  max_position: 5
  max_daily_loss: 1500
  daily_reset_time: "16:00 ET"

For a setup that uses both (copying from a Rithmic master account to one Rithmic follower and one ProjectX follower) the destinations are configured independently and the sizing rules apply per follower. There's no relay layer to manage and no manual reconciliation.

How to choose

The decision in 2026 is rarely "one or the other": it's "which one first, and when do I add the second." Defaults that have held up well:

1. If you're starting on Apex, Bulenox, Tradeify, or any of the long-tail futures prop firms: you're on Rithmic by default. Don't fight it. Rithmic is the path of least resistance and the platform tooling around it is mature.
2. If you're starting on TopStep: you're on ProjectX. Use TopStepX or a copy trader that handles ProjectX (most modern services do). Don't try to route TopStep through Rithmic: that's not how the firm structures the connection.
3. If you scalp tightly (under 5 ticks of edge per trade): Rithmic. The p99 latency advantage is real, and the gap shows up most in the trades that matter most to your edge.
4. If you swing-trade or hold positions for hours: Either is fine. Pick whichever your prop firm uses and stop overthinking it.
5. If you run a multi-firm stable: Use a copy trading service that handles both natively. The configuration overhead disappears, and adding a new firm becomes a 5-minute task.

For technical depth on the protocols themselves, Rithmic publishes official R | API+ documentation, and CME Group's data licensing rules, which apply to both feeds, are explained on the CME non-professional subscriber page. Most issues that look like "feed bugs" are actually licensing-tier mismatches, and the CME documentation is the authoritative source on what tier you fit in.

Frequently asked questions