How to Compare Prop Firm Rules with Claude (Step-by-Step)

The fastest way to compare prop firms is to stop reading their marketing pages and compare their rulebooks side by side. Claude is very good at that job. You paste in each firm's rules, it pulls out the same fields for every firm into one table, quotes the exact sentence behind each number, and flags the clauses that quietly decide whether you get paid. This tutorial walks through the whole workflow with prompts you can copy, from gathering the rules to testing them against how you actually trade.

TL;DR

Collect each firm's current rules, FAQ, and terms. Put them in a Claude Project so you only explain the job once. Run one extraction prompt per firm that forces a fixed set of fields plus an exact quote for each, then a comparison prompt that lines the firms up and does the drawdown math. Finish by describing your own trading style and asking which rules you'd break. Claude does the reading and organizing; you check every number against the source. It can't tell you whether a firm pays, so check payout proof separately. The free Claude Cowork course teaches the same approach for any document-heavy decision.

Why prop firm comparison is harder than it looks

Two firms can both advertise "$25K account, 90% split, from $99" and be completely different products. The differences hide in a handful of rules:

• Drawdown type. A static drawdown is a fixed floor that never moves. A trailing drawdown follows your highest balance or equity upward, so giving back open profit shrinks your room. Same percentage, very different account.
• Consistency rules. Some firms cap how much of your total profit can come from your best day. Pass the profit target with one big day and you may still be blocked from a payout.
• Payout terms. Minimum trading days, payout caps, waiting periods, and whether a payout resets or locks your drawdown floor.
• Trading restrictions. News trading, weekend holding, copy trading, hedging across accounts, maximum leverage per market.

The rules are spread across a pricing page, an FAQ, a help center, and a terms of service, and they're often worded differently in each place. That's exactly the kind of tedious cross-referencing an AI handles well, as long as you make it show its sources.

If you just want the numbers already organized, PerpsMatch's prop firm comparison (opens in a new tab) lines up hundreds of perps evaluation plans by price, profit target, max drawdown, and split, with each firm's rules checked against its own docs. The workflow below is for going deeper on your shortlist, or for comparing any firm, in any market, that a comparison site doesn't cover.

Step 1: Gather the rules from the source

Don't ask Claude what a firm's rules are. Its training data is out of date, and prop firms change their rules constantly. It may give you an old version or invent one. Always give it the current text.

For each firm on your shortlist, copy or save as PDF:

1. The challenge or pricing page for the exact plan you're considering
2. The rules or FAQ page
3. The payout policy
4. The terms of service (the part about prohibited strategies and account termination)

Note the date you copied it. If the firm has a help center, search it for "drawdown," "consistency," "payout," and "prohibited" and grab those articles too.

Three firms is a good number for one pass. More than five and you'll spend more time checking than comparing.

Step 2: Set up a Claude Project

A Claude Project keeps your instructions and files together, so every chat starts with the context already loaded. (New to Projects? See Claude Projects explained.)

Create a project called "Prop firm research," upload the documents from Step 1, and paste this into the project instructions:

You help me compare prop trading firms using only the documents I provide.

Rules:
- Never use outside knowledge about any firm. If a detail is not in the
  documents, write "Not stated" — do not guess.
- For every rule you report, include the exact sentence from the source
  and which document it came from.
- If two documents from the same firm contradict each other, flag it as
  "CONFLICT" and quote both.
- Show your math step by step whenever you calculate anything.
- Do not recommend a firm. Lay out the facts and the tradeoffs.

The "Not stated" and "CONFLICT" rules do most of the work. They turn Claude's main weakness, confidently filling gaps, into something you can see.

Step 3: Extract each firm into the same fields

Run this once per firm, in its own chat, so the firms don't blur together:

Using only the documents for [FIRM NAME], fill in this table for the
[$25K 1-Step] plan. One row per field. Columns: Field | Value | Exact quote | Source doc.

Fields:
1. Evaluation price
2. Steps / phases
3. Profit target (each phase)
4. Daily loss limit — and is it based on balance or equity?
5. Max drawdown — static, trailing intraday, end-of-day trailing, or hybrid?
6. Does the trailing drawdown stop (lock) at the starting balance?
7. Minimum trading days
8. Time limit
9. Consistency rule (exact percentage and how "best day" is defined)
10. Max leverage by market
11. News trading allowed?
12. Weekend holding allowed?
13. Prohibited strategies
14. Profit split (and any paid upgrade)
15. Payout frequency and waiting period
16. Payout cap or minimum
17. What happens to the drawdown floor after a payout?
18. Reset or refund policy

After the table, list anything unusual that could stop a trader who
hit the profit target from getting paid.

Then check it. Spot-check at least the drawdown, consistency, and payout rows against the quotes. This takes five minutes and catches the occasional misread. If a quote doesn't say what the value column says, the value is wrong.

Step 4: Compare the firms side by side

Once you have a verified table for each firm, paste them into one chat:

Here are the verified rule tables for [Firm A], [Firm B], and [Firm C].
Build one comparison table with the firms as columns.

Then add these calculated rows, showing your math:
- P/D ratio: profit target ÷ max drawdown (lower = easier to pass)
- Cost per $1,000 of max drawdown: price ÷ (max drawdown in dollars / 1,000)
- Dollars of room on day one: starting balance minus the drawdown floor

Finally, list the 3 biggest practical differences between these firms
for someone who trades [describe in one line, e.g. "BTC and ETH perps,
2-4 trades a day, sometimes holds over the weekend"].

The P/D ratio is a quick way to see how hard a plan is. A 10% target with a 5% max drawdown (P/D of 2) is much harder than an 8% target with a 10% drawdown (P/D of 0.8). Comparison sites like PerpsMatch use the same idea to sort plans by how easy they are to pass.

Step 5: Test the rules against how you actually trade

This is the step most people skip and the one that matters most. A rule only hurts you if your style runs into it.

Run a trailing drawdown scenario

Trailing drawdowns are where traders get surprised. Make Claude walk through a real sequence:

Using [Firm B]'s drawdown rule, walk me through this sequence on a
$100,000 account, step by step, showing the drawdown floor and my
remaining buffer after each step:

1. Start of account
2. An open trade goes to +$5,000 unrealized
3. The trade reverses and I close it at +$500
4. Next day I lose $2,000

Then do the same sequence under [Firm A]'s rule and compare the
buffers.

With an 8% trailing drawdown on equity, that open +$5,000 moves your floor from $92,000 up to $97,000. Close at +$500 and your buffer has shrunk from $8,000 to $3,500, even though you're in profit. Under an 8% static drawdown your buffer would be $8,500. Seeing that in numbers changes which firm you pick. For a deeper explanation of the four drawdown types, PerpsMatch has a clear guide on static vs trailing drawdown (opens in a new tab).

Verify the math. Claude is good at this kind of arithmetic but not perfect, especially across many steps. Check the final numbers with a calculator or a drawdown calculator.

Check the consistency rule against your best day

If a firm has a consistency rule, paste in your real (or demo) daily P&L:

Here are my daily P&L results from the last month: [paste].
Under [Firm C]'s consistency rule, would I be eligible for a payout?
If not, how much more total net profit would I need, assuming no new
best day? Show the math.

The math is simple. If your best day must be no more than 30% of total profit and your best day was $600, you need at least $2,000 total. Claude will usually get this right, but a dedicated consistency rule calculator (opens in a new tab) is a quick way to double-check.

Ask what you'd break

Here is how I trade: [instruments, typical position size, leverage,
how long I hold, whether I trade news, whether I hold over weekends,
whether I use bots or copy trading].

For each firm, list every rule my style would break or come close to
breaking, with the exact quote. Rank the firms from fewest conflicts
to most.

Step 6: Check the one thing Claude can't

None of this tells you whether a firm actually pays. Claude can only read what a firm says about itself. A perfect rulebook from a firm that runs out of money is worth nothing.

For that you need evidence from outside the firm: payouts you can see on a block explorer from a public wallet or contract, recent payout reports from real traders, and a known company or backer. For perps firms, PerpsMatch follows published payout wallets and marks firms that only report their own totals, which is a fair way to handle a market where most firms are under two years old. Whatever you use, do this step before you buy, not after you pass.

Step 7: Save the workflow so you can reuse it

Prop firms update their rules often, and you'll want to rerun this when they do. Turn the Step 3 prompt and the Step 2 instructions into a reusable Claude Skill so you can say "extract this firm's rules" and get the same table every time. How to create a Claude Skill walks through it.

Where this workflow goes wrong

• Stale or partial rules. If you paste last month's FAQ, you get last month's answer. Recopy before every decision.
• Missing documents. A rule that only lives in a help-center article won't appear if you didn't include it. "Not stated" doesn't mean "not a rule."
• Multi-step math. Long drawdown sequences are where small arithmetic errors creep in. Verify the final numbers.
• Ambiguous wording. When a rule could be read two ways, Claude will pick one. Ask the firm's support team directly and save their answer.
• No predictive power. Claude can tell you which rules fit your style. It can't tell you whether you'll pass, or which firm will still be around next year.

Frequently asked questions

Can Claude compare prop firms for me?

Yes, if you give it each firm's current rules. Claude can extract the same fields from every firm into one table, quote the source for each rule, calculate things like P/D ratio and drawdown buffers, and flag rules your trading style would break. Don't rely on what it knows from training, because prop firm rules change often.

What is the difference between static and trailing drawdown?

A static drawdown is a fixed loss floor set below your starting balance that never moves. A trailing drawdown rises with your highest balance or equity, so giving back open profit reduces your remaining room. Most trailing drawdowns stop rising once the floor reaches your starting balance.

What is a prop firm consistency rule?

A consistency rule limits how much of your total profit can come from a single day, for example no more than 30 percent. If your best day is too large a share, you need more total profit before you can request a payout, even if you already hit the target.

What should I compare when choosing a prop firm?

Compare the price, profit target, daily loss limit, max drawdown type, consistency rule, trading restrictions like news and weekend holding, profit split, and payout terms. Then check payout proof separately, since the rules only matter if the firm pays.

Is it safe to use AI to pick a prop firm?

It's safe for reading and organizing rules, as long as you give it current documents and check its work against the quotes. AI can misread a clause or make an arithmetic mistake, and it can't verify whether a firm pays traders. Treat its output as a research summary, not a recommendation.

Learn the workflow, then use it everywhere

This is the same method you'd use to compare insurance policies, vendor contracts, or software plans: give Claude the source documents, force a fixed structure, demand quotes, and verify. The free Claude Cowork course teaches it step by step in plain English. If you're on the other side of the table and thinking about launching a firm, read how to start a prop firm.