CS2 Case Drop Rates Guide (2026)

Did you know, your odds of unboxing a knife from a CS2 case are 0.26%? Yup! That is roughly 1 in 385 openings. And, a StatTrak knife? One in 3,850!

These are not estimates or community guesses, either. They are the exact numbers Valve was forced to publish in 2017 when China's Ministry of Culture required every loot-box game to disclose its drop rates publicly.

And, the numbers have not changed since then. Not when CSGO became CS2. Not when new cases like the Kilowatt and Gallery dropped. The same drop rates govern every weapon case, and Valve has never adjusted them.

But what are the exact drop rates? This guide breaks down every one across weapon cases, souvenir packages, and sticker capsules, plus the math behind why opening cases is almost always a loss!

These are Valve's official drop rates for every CS2 weapon case, from the original CS:GO Weapon Case all the way through the Kilowatt Case and beyond. They follow a rough "5x rule" where each rarity tier is approximately five times rarer than the one below it.

These rates are identical across every weapon case in the game. For example, opening a Kilowatt Case, a Revolution Case, a Dreams & Nightmares Case, or a vintage Bravo Case all produce the same probability distribution. The only thing that changes between cases is which specific skins sit in each rarity tier, and therefore how much money those skins are worth on the Steam Community Market.

The rates were disclosed through Valve's Chinese publishing partner, Perfect World, after China's Ministry of Culture mandated loot-box transparency. Community data scientists have since verified these numbers using aggregate datasets of millions of case openings, CS2 case-opening simulators, algorithms, and manual testing.

Every eligible item you unbox from a weapon case has an independent 10% chance of being StatTrak. That 10% applies on top of the rarity roll. So when you get a Mil-Spec skin (79.92% base chance), there's a 10% chance that Mil-Spec skin will also be StatTrak. Simple multiplication gives you the combined odds.

StatTrak Mil-Spec: 7.99%. StatTrak Restricted: 1.60%. StatTrak Classified: 0.32%. StatTrak Covert: 0.064%. StatTrak Knife or Gloves: 0.026%, which works out to roughly 1 in 3,850 case openings.

That means if you opened a case every single day, it would take you over ten years on average to hit a StatTrak knife! 

The StatTrak modifier is binary and independent. It doesn't affect which skin within a rarity tier you receive, and it doesn't interact with wear values. A StatTrak Factory New Covert is the mathematical equivalent of hitting three separate low-probability rolls in sequence.

Every skin you unbox gets a random float value between 0.00 and 1.00 that determines its wear condition. The float distribution is not uniform. Factory New skins are significantly rarer than Field-Tested ones, which matters a lot for the market value of what you unbox.

The approximate wear distribution from case openings breaks down like this: Factory New (0.00 to 0.07 float) appears in roughly 3% of drops. Minimal Wear (0.07 to 0.15) shows up about 24% of the time. Field-Tested (0.15 to 0.38) is the most common at around 33%. Well-Worn (0.38 to 0.45) is about 24%, and Battle-Scarred (0.45 to 1.00) rounds it out at roughly 16%.

Here is where it gets painful. Combine the 0.64% Covert drop rate with the ~3% Factory New distribution, and you get about a 0.019% chance of unboxing a Factory New Covert skin. That is roughly 1 in 5,200 openings. A Factory New StatTrak Covert? You are looking at 1 in 52,000. At $2.50 per key, that is an expected cost of $130,000 before you see one.

Individual skins also have float range caps. Not every skin can roll Factory New. The AWP Asiimov, for example, only exists in Field-Tested, Well-Worn, and Battle-Scarred conditions because its float range starts at 0.18. These caps are set per skin by Valve and do not change.

Souvenir packages play by different rules than weapon cases. They contain six rarity tiers instead of five (adding Consumer and Industrial grades at the bottom), they do not require a key to open, and they cannot contain knives or gloves. The tradeoff is that Souvenir Covert skins are absurdly rare.

Souvenir packages follow the same 5x multiplier pattern but across six tiers. Consumer Grade drops at 80%. Industrial Grade at 16%. Mil-Spec at 3.2%. Restricted at 0.64%. Classified at 0.128%. And Covert? A stunning 0.0256%, which works out to about 1 in 3,906 packages.

That makes a Souvenir Covert nearly 40 times rarer than a regular weapon case Covert and roughly 10 times rarer than a knife. The Souvenir AWP Dragon Lore from the Cobblestone Collection is the poster child for this rarity. If you tried to unbox one from souvenir packages alone at $10 per package, the expected cost would land somewhere around $39,000, and that is before you account for the specific skin you want within the Covert pool.

Souvenir skins carry gold tournament stickers from the specific Major match they are tied to, which adds collector value. But opening them for profit is even worse math than weapon cases. If you want a specific Souvenir skin, buy it on the market. Every time.

Sticker capsules from CS2 Majors use a simplified four-tier system. High Grade (paper stickers) drop at 80%. Remarkable (holo stickers) at 16%. Exotic (foil stickers) at 3.2%. Extraordinary (gold stickers) at roughly 0.64%, which is about 1 in 156 capsules.

Gold stickers are the big-ticket items. A gold sticker from the right player or team at the right Major can be worth hundreds of dollars, particularly if that player retires or changes teams shortly after. The odds of pulling a specific gold sticker depend on how many stickers are in the capsule, but the base 0.64% chance of hitting gold at all keeps these firmly in the "lottery ticket" category.

Autograph capsules follow the same distribution. Whether you are opening a team sticker capsule or a player autograph capsule, the 80/16/3.2/0.64 split holds. The only variable is the number of stickers in the pool and their individual market values.