Good Servants to Grail for their Noble Phantasms

I have most of the Arts Single Target NP Sabers and, I’m wondering if I should grail my level 80 Hokusai to level 90 while the event is still going on so I can get that 2x exp bonus for her! I’m asking because I currently have all the NA Arts Single Target NP Sabers except for Dioscuri, and Nero’s Bride form. I know we are getting two more of these sabers in the form of this year’s Gudaguda event with the 4 stars Hajime Saito and The Trung Sisters in about 2 years’ time. Now out of all the servants that I don’t have or the ones who are not yet available, Dioscuri and the Trung Sisters are who I want the most. Down below is a list of all my current Arts Single Target NP Sabers and their levels and Np levels.

They are:
Katsushika Hokusai Level 80/80 Np 5
Yagyu Munenor Level 72/80 Np 4
Lancelot Level 80/80 Np4
Queen Medb Summer Level 80/80 Np 1
Beni-Enma Level 90/90 Np 1


If Hokusai Saber is a favorite, grail and enjoy; if not, there is no pressing need to grail her, especially if you’re calling it quits at 90. Not at all necessary for gameplay.


What Gou said.

To add, no one is essential to grail. Sabers get average benefit from grails, but ST Saber is a very small niche for your grails.

Just grail superfavs. They’re a limited resource so don’t rush to use them.


Grail your favourites, grails are rather limited so think it through before grailing randomly.

Personally the only ones worth grailing for gameplay reasons are universal farmers (Dantes, Space Ishtar etc.) that you’ll bring pretty much everywhere. For these servants, grails/gold fous/higher NP levels will help them hit higher damage/refund thresholds for farming.


As the others say, you should consider this the other way around:

I love that servant, I want to use him/her everywhere until the day I die but damn, that damage is abysmal.

If you only need some more ATK stat, try levelling the CE they wear instead of grailing them. Getting them around lvl 50 is a good compromise between EXP/QP costs and the ATK gain in my experience. Obviously you’ll want to favour full ATK CEs for it to work best.
CE EXP can be hard to come by early in the game, so select your candidate CEs carefully.


On a related topic, this might be common knowledge for veteran players, but did you mention 5x super and great success rates during 5th anniversary? I remember reading base super (or whatever the 2x exp one is called) chance being 10% and great chance being 3%, and math based on that suggests that even 5x super and great rate gives less expected xp compared to 2x exp for event/welfare servants, something like 1.8x exp vs 2.32 exp. Am I right in understanding that 2x exp for Hokusai currently is better than waiting for 5th anniversary if someone wants to grail/level her up?

1 Like

I think grailing Arash is also a good investment. When grailed and given 2k Fou, he can clear decently chunky waves (depending on CE and buffs) and bring in a support, which can be useful for 6 slotting some 90+ nodes (that is, clearing them while every servant is equipped with an event CE) in the future.

1 Like

grailing for stats its not worth it unless you dont care for your servants.

1 Like

The average with both bonuses considered at anni is 1.7x exp*. Just level her now if her bo us is still active, she has 2x exp base atm which is then multiplied by successes.

As for anni:

Thanks to half AP with 5 star ember nodes and that 1.7x exp, it should be ~200AP to max an SSR. About double that to go lv1-100.

JP didn’t have 5 star ember nodes at 5th, we’re gettimg them early. Anni is normally 3x success, on JP 5th had 5x success as a special landmark.

*Head maths for this:

  • 2x success has a 50% chance, that makes the average benefit +50% or 1.5x (2x exp is +100%, X 0.5)
  • 3x success has a 10% success rate, that makes the average benefit +20% or 1.2x (3x exp is +200%, X 0.1)

Add those together for a 1.7x average benefit.

Grails are a very small boost to stats comparing to everything else.
You can think of it like this (It’s NOT an absolute rule for everyone but at least %80 of servants follow this logic)
It’s never a waste to grail a servant if you ask me but the game design to favor one servant over everyone else for limited things like fou stuff(fou prints, fou cards etc.)+grails

2 weeks ago I did math for someone asking whether they should grail their Hokusai to lv90 or not
Turned out levelling a CE gives her bigger damage boost than grailing
So keep that in mind before you decide to spend the rarest resource in the game (especially that will be the case for the rest of servants you’ve listed too)

1 Like

Fun fact: “Angry Mango” is probably the most correct answer to OP’s question.

1 Like

The only servants you should grail for stats are the universal farmers. While the boost is small, it’s on the base attack, which means it acts multiplicatively with every other buff you pile on them during farming.

For a quick example, Lv90, 1k/1k, NP1 Dantes, with full DSS buffs (not counting MC), deals between 83k-101k damage.
At Lv100, with all the same buffs, you’re looking at 90k-110k. That’s an average of 9% damage increase, which isn’t huge, but it can theoretically let you clutch a damage threshold.

On the other hand, Lv120 grailing does provide quite the buff. As of now, my hard-refill DSS Spishtar, at Lv100, does around 220k damage on the final wave. Once I get her up to Lv120 in the distant future, that should be around the 250k mark, which definitely isn’t insignificant, and it’s going to pay even bigger dividends with DCS setups.

That said, if you really want to raise a farming servant’s damage output, NP levels are unfortunately still the most impactful way to do it by far.

Not true unless you’ve proven with math that attack gain from grailing makes given setup consistent in a node that you’re going to run 300+ times (which would be a case only for raids, lottos, maybe HQs)

Like sure, Dantes gets 7k damage buff, but do you need it if the given w3 has boss with 82k HP and refund is consistent across all three waves?

IMO given the experience I have with the community, if anybody asks about “gameplay grails” should hold back from grailing because they clearly don’t understand how little gameplay value grailing has. And even if it has, how much does it really cost and how many benefits does it bring? These are the real questions


To be fair, I want my universal farmer to be universal. I do not really care if it works for a node I will run 300 times. I want it to work for every conceivable node. That being said, Rin is my favorite character, so you know, it was not like I sacrificed anything by simping to the nth degree. If it doesn’t work in the set up despite the highest possible investment, it just wasn’t meant to be. I will live.

1 Like

Erm, alright?
But we’re talking about OP who asked who they have to grail for so-called “gameplay” and they didn’t mention they have crush on any of these servants
If the waifu factor was in, I wouldn’t say a word because waifu>meta

So with all honesty regardless what do you waifu, whether its Spacebar and Mata Hari, as long as you waifu them and love dearly, its absolutely alright to grail

That is great and all, but I was responding to your response to someone else talking about universal farmers. You said it was not true, but I believe it is true as the word universal typically means something that is not defined as “a specific mode you’re going to run 300+ times.” Hope that clears it up, Ivory.

Their point is “do the grails actually push you over the damage/refund line, or is it more likely to simply be winmoar?”, and they’re suggesting winmoar.

If you already have NP5, then unless the Servant has low atk/steroids, grails are not gonna make the difference. In the former case, -120- grails are going to make the difference and that NP5 probably indicates waifubando status.


I cannot recall the actual math, but I believe grailing can simulate a ghetto np2 (it might be grails plus golden fous). The advice here is usual awful as people suggest an np1 Dantes or Space Ishtar is ideal for farming, and that might only be true if you are willing to pimp them out, and then it usually comes at a cost on the CE or mystic code.

It all depends on the situation. I agree that np5 is not really necessary, I am just saying that there is definitely some gray space where grails can make the difference depending on np level and what servant we are talking about. That is not really debatable.

1 Like

You’re correct about grails, Fous, etc. being able to bridge or nearly bridge some NP-level thresholds, but the general point is that these are specific-use cases. You don’t grail just for gameplay unless you have that kind of goal in mind, and for most players it is not worth it to grail a Servant they don’t truly like.

I don’t believe I’ve seen many people here trying to sell NP1 universal farmers except arguably Arjuna Alter in a future context of DKSO. NP2-3 for the likes of Space Ishtar and the other top Castoria farmers that can often run with your CE of choice is not unrealistic if universal farming is the goal.