Cooking buffs: +50% max HP, and the grade does not matter
✍️ Alice・愛麗沙
Whatever you eat, the health bonus is identical: max HP +50% for 15 minutes, with no dependence on the dish’s cooking level or grade. That single line answers half of “should I eat the expensive one” — if you want health, eat a boiled egg at 50 gold. What actually varies by recipe is all-stat: the shared tiers print +85 / +155 / +185, and the dishes at the top of the ladder cost tens of thousands of gold in ingredients. Work the margin and the exchange rate is about 592 gold per point of all-stat — that is what the extra spend buys, not health. Then there is the uncomfortable part: a life membership hands out three lunch boxes a day for free, and the Korean community’s verdict is that the membership box beats a tier-3 dish outright. So the correct play on this skill is “cheapest dish normally, lunch box when it matters”, not “level cooking and cook for yourself”. Three things we could not find and will not invent: how long the all-stat portion lasts, whether it survives death or logout, and whether two dishes stack.
One sentence removes an entire levelling anxiety
Start with the cheap part.
The Korean wiki states the effect plainly: eating any dish gives max HP +50% for 15 minutes, and that effect has nothing to do with the dish’s grade. The footnote goes out of its way to say “independent of level”, which tells you how counter-intuitive the rule is — it needed writing down explicitly.
So the question “do I have to level cooking to get a good buff” simply disappears in the health column. A boiled egg and a 50,000-gold fish steak return exactly the same health. And the boiled egg costs 50 gold, one empty bottle and one egg.
Every difference lives in the other column.
All-stat is the thing you are actually buying
Each recipe separately prints a fixed all-stat number. It does not scale with your level; it is hardcoded dish by dish:
| Category | All-stat | Approximate ingredient cost |
|---|---|---|
| Boiled egg and its kind | Effectively none | 50 |
| Shared tier 1 (white fish meuniere, bouillabaisse, mackerel and salmon steak, and so on) | +85 | 18,000–50,400 |
| Shared tier 2 | +155 | Higher |
| Shared tier 3 | +185 | Higher |
| Epic feast | +175 | Higher |
| High-end dessert (cooking Lv.25, station Lv.6) | +185, plus 5,200 HP every 10 seconds for 10 minutes | Higher |
Subtract the cheapest from the dearest and the marginal ledger reads:
50,400 − 50 = 50,350 gold, buying +85 all-stat (a tier-1 shared dish) which is about 592 gold per point of all-stat
Is that unit price expensive? It depends what you hold it against. Take the gold-shop gem box as a benchmark: 40,000 gold each, and the endpoint of that chain — the star prism — is a permanent slot giving both all-stat and max HP. The dish’s +85 lasts one fight; the gem’s bonus stays on the account. Same money: fifteen minutes on one side, indefinitely on the other.
That is what the Korean community means when it says you cannot cook casually. Not that the ingredients are unaffordable — that every meal competes with a permanent purchase. The ingredient prices show why: asparagus 10,000, peas 10,000, tomato 6,800, lemon 3,000 with a cap of thirty a week per shop. Meanwhile a gold ingot cashes out at 5,000. A single ingredient slot on a high-tier recipe costs more than two ingots.
The campfire is where cooking is actually fought
If cooking were only for feeding yourself, the ledger above would end the article. It does not end there, because there is a second set of effects: campfire sharing.
Put a shared dish on a campfire and the effect is handed to every party member:
- Tier 1: attack +120, defence +120, 10 minutes
- Tier 2: attack +170, defence +170, 10 minutes
- Tier 3: attack +220, defence +220, 10 minutes
- Epic feast (one version): attack +500, 10 minutes, plus an area burn
- Epic feast (the other): attack +800, 20 minutes, plus life EXP +10% and non-combat movement speed +10%
The arithmetic here is completely different, because the denominator becomes party size. A tier-3 dish eaten alone is +220. Shared across four people it is 4 × 220 = 880 points of attack and defence delivered, for a cost paid once. The attack +800 / 20-minute epic feast is more extreme still — it is one of the few things in this game that gets a life-skill player named in a group-content roster.
So cooking’s value splits into two separate questions, and most guides run them together:
- Eating alone: only the cheapest dish is worth making, because the health column is identical anyway.
- Lighting a campfire for the party: this is why the high-tier recipes exist at all, and the return scales directly with party size and content difficulty.
The third exit: hand the dish over and take goods
There is one more way out besides eating and sharing — using a cooked dish as the entry ticket for an exchange.
| Dish | Cooking level | Cooking station | NPC (town) | Get | Cap |
|---|---|---|---|---|---|
| 旅行者點心 | Lv.1 | No station needed | 貓商人（地下城3區） | 鹽 ×2 | 1/day |
| 馬鈴薯沙拉 | Lv.1 | Lv.1 | 阿蘭雯（杜巴頓） | 絕招秘藥 ×1 | 1/day |
| 蘋果汁 | Lv.1 | Lv.1 | 麗莎（堤爾克那） | 高級鍊金術再燃燒催化劑 ×1 | 1/day |
| 美乃滋炒肉 | Lv.1 | Lv.1 | 雷納德（堤爾克那） | 暴擊秘藥 ×1 | 1/day |
| 擠著吃的點心 | Lv.3 | Lv.1 | 貓商人（地下城3區） | 愛心幣 ×1 | 3/day |
| 炒蔬菜 | Lv.5 | Lv.2 | 艾爾文（杜加德走廊） | 高級木材 ×4 | 2/day |
| 蘋果舒芙蕾 | Lv.5 | Lv.2 | 潔米（杜巴頓） | 高級布料 ×4 | 2/day |
| 瑞可塔起司沙拉 | Lv.5 | Lv.2 | 傑羅姆（杜巴頓） | 絲綢 ×4 | 2/day |
| 牛奶燉豬肉 | Lv.5 | Lv.2 | 格莉娜（杜巴頓） | 格莉娜手工蛋糕 ×1 | 1/day |
| 蘋果舒芙蕾 | Lv.5 | Lv.2 | 格莉娜（杜巴頓） | 食物製作台設計圖(3級) ×1 | 1/day |
| 蘋果鮮奶油蛋糕 | Lv.10 | Lv.3 | 潔米（杜巴頓） | 高級布料+ ×4 | 1/day |
| 奶油醬牛排 | Lv.10 | Lv.3 | 傑羅姆（杜巴頓） | 高級絲綢 ×4 | 1/day |
| 蒜香橄欖油義大利麵 | Lv.10 | Lv.3 | 基利安（庫漢） | 四葉草 ×8 | 1/day |
| 冰草莓飲 | Lv.10 | Lv.3 | 凱雅拉（庫漢） | 高級暴擊秘藥 ×1 | 1/day |
| 馬鈴薯湯 | Lv.10 | Lv.3 | 卡琳（冰霜峽谷(雪山救助隊)） | 寵物飼料 ×10 | 1/day |
| 沙威瑪 | Lv.13 | Lv.4 | 休馬斯（萊爾特丘陵(龍之遺跡)） | 卓越繃帶 ×5 | 3/day |
| 豆乳防風草蛋糕 | Lv.13 | Lv.4 | 康格（班克爾） | 卓越恢復藥水 ×6 | 1/day |
| 鱸魚辣魚湯 | Lv.20 | Lv.4 | 艾頓（班克爾） | 銀合金錠 ×10 | 1/week |
| 蒸蛤蜊 | Recipe unverified | — | 特蕾西（杜加德走廊） | 特蕾西的原木音樂盒 ×1 | 1/day |
Only live TW routes whose entry ticket is a COOKED dish; barter-only foods (the wholemeal bread, the handmade cake) cost no cooking levels and are not listed. Cooking and station levels are KR player records. Full barter board:Barter
Read the gate columns rather than the individual items. The entry tickets cluster at the low-level dishes: cooking Lv.1 / Lv.5 / Lv.10 with a food station at Lv.1–3 opens most of the list.
Which gives cooking a comfortable property. The reason to level it for the exchange lane runs out around Lv.10. The skill caps at Lv.35 — not 37, despite what gets repeated — and the twenty-five levels above ten exist for buffs, not for trades. And the buff side, as the ledger above shows, is priced absurdly.
The membership lunch box, which flattens the whole line
The last section is the Korean community’s own conclusion, and it stings.
A life membership hands out three deluxe lunch boxes every day. The players’ verdict: the membership box outperforms a tier-3 dish. So actual play collapses into “boiled egg normally, lunch box when it matters”, and nobody cooks the expensive middle of the ladder at all.
The Taiwanese storefront prices that membership at NT$107 a month, which is the figure to hold against a weekly ingredient bill running into six figures of gold. But the point is not the subscription price — it is what belongs in the comparison set. Measured against not eating, cooking is obviously worth it. Measured against three free lunch boxes a day, the answer changes.
Three things we could not find
Listed, because each of them changes how you would use these buffs:
- How long the all-stat portion lasts. Only the +50% health carries an explicit 15 minutes. Whether all-stat runs on the same timer has no source in either region.
- Whether it survives death or logout. No data at all.
- Whether two dishes stack. The claim that same-family buffs do not stack circulates widely and has no traceable origin.
One regional conflict also deserves flagging. A Taiwanese newbie guide states that cooking can raise health by up to 150%, which does not reconcile with the Korean +50%. Neither side could be checked word for word against an official text, so this article uses the Korean version — the one with a footnote and explicit wording — and leaves the other claim visible rather than deleting it.
Where this lands
Cooking is a skill with three exits, and most players only look at the most expensive one. Exit one is eating it yourself, where health is identical so the cheapest dish wins. Exit two is campfire sharing, the only genuinely profitable use of high-tier recipes, and it requires a party. Exit three is trading, whose gates run out around Lv.10.
Do: level far enough to open the trade list, then stop. Always eat the cheapest dish when eating alone. Buy expensive ingredients only when you intend to light a campfire, and buy them per run. Ignore: the goal of maxing cooking to Lv.35 as an end in itself — if you are already running content that needs shared buffs, you will know it without being told.
Two things would overturn this page.
If the health bonus really is +150% rather than +50%, “eat the cheapest” still holds — it is grade-independent either way — but cooking’s standing as a sustain tool rises sharply. That cell is worth someone measuring in-client once.
And if the all-stat portion lasts far longer than 15 minutes — a whole dungeon run, say — the cost per minute of a high-tier dish drops hard and the marginal ledger in section three needs recomputing. Nobody has published that number.
