HARMA Card Upgrades and Fusion Guide
How HARMA X/Y card upgrades work, when to fuse duplicates, and which upgrades are worth the gold in Early Access.
HARMA gives every card an upgrade choice, and the choice is not cosmetic. At an altar or upgrade screen you pick an X or Y branch. X usually doubles down on the card's existing job: more damage, more block, a bigger number tied to attunement successes. Y usually adds a new job: extra draw, a secondary effect, or a payoff that connects to another system. Picking the wrong branch turns a good card into a card you never play. This guide explains the decision, not a fixed tier list, because the right branch depends on your deck and your artifact spread. If you are still choosing a lane, the best builds overview→ explains which upgrades matter for each affinity.
The X Branch: Make the Card Better at What It Already Does
X upgrades are the safe pick when the card is already doing the work you need. If your main attack is the only thing killing enemies, upgrading its damage is correct even if the upgrade is boring. X branches often scale with attunement results: the two-success and three-success numbers go up, or the card gains a bonus when a specific artifact succeeds.
Take X when:
- The card is your primary damage or block source.
- You need a breakpoint, not a new engine.
- Your deck is already coherent and you do not want another decision point.
- The Y branch adds a resource you are not set up to use.
The trap is upgrading a card you play twice a run because the X number looks big. Upgrade frequency matters more than the size of the number. A card you play every combat is a better upgrade target than a situational finisher.
The Y Branch: Add a Second Job
Y upgrades are for cards that are already good enough at their first job. A damage card that also draws a card becomes a consistency engine. A block card that also applies a status becomes a bridge between your defense and your payoff. Y branches are stronger in decks with room to exploit the extra effect, and weaker in decks that need raw numbers.
Take Y when:
- The base card is cheap and you will play it often.
- The extra effect fills a gap in your deck (draw, status, energy, artifact manipulation).
- You have a payoff that cares about the added effect.
- You are not relying on this card to hit a exact damage breakpoint.
Count how often a card leaves your hand. If it sits there for two fights, it is not the right upgrade target.
Fusion: Turn Duplicates into Upgrades
Duplicate cards can be fused. Fusion combines them into a stronger version and then offers another X/Y choice. This is one of the best uses of duplicates because it removes a redundant draw and adds power at the same time. The catch is that you lose the second copy, so fusion is bad when the duplicate is a core card you want to draw every turn.
Fuse when:
- The duplicate is a card you only need once per fight.
- You have a second copy of a card that does not stack well.
- You want to consolidate power into one card and free a deck slot.
Do not fuse your only copy of a key card. Do not fuse two cards just because the game offers it. A duplicate is still a card; a fused card is a commitment.
Upgrades vs. Removal
Removal and upgrades compete for the same gold. The rule of thumb: remove cards that do nothing, upgrade cards that do everything. If a card is merely mediocre, removal is often better because it improves every future draw. If a card is central to your plan, upgrade it and keep it.
A 12-card deck is an achievement, but it is also a real strategy. Fewer cards means more consistent draws and more frequent access to your best card. Do not remove so aggressively that you run out of playables, though. The deck-building mistakes→ page covers the balance.
Upgrade Priority by Card Role
| Role | Preferred branch | Why |
|---|---|---|
| Main attack | X | Hitting a damage breakpoint matters more than utility |
| Cheap block | Y | Extra draw or status turns a defensive card into a bridge |
| Setup card | Y | Adds consistency or a second payoff |
| Finisher | X | You need the maximum tier to be worth the setup |
| Duplicate | Fuse | Consolidates power and frees a slot |
| Dead starter | Remove | No upgrade saves a card you never play |
The Alchemist Is Not a Free Upgrade
The Alchemist offers multiple services, and the achievement for using four different services is a hint that the shop is a system, not a button. Check what each service costs before you commit. Sometimes a card change is better than a raw upgrade, and sometimes saving the gold for a removal is better than either. The economy guide→ explains the full menu and the opportunity costs.
When to Stop Upgrading
Stop when your deck has a clear plan and every card either advances it or keeps you alive. Extra upgrades after that point are usually vanity. A lean deck with two or three key upgrades beats a bloated deck where every card has been touched once. If you cannot explain why a card is in the deck, remove it instead of upgrading it.

FAQ
Should I always take X on attacks? Usually, but not always. If the attack already kills the targets you need and the Y branch adds draw or a status your build wants, Y can be better.
Is fusion ever bad? Yes. Fusing away a duplicate of your best card can reduce consistency. Fuse cards you only need once per fight.
Can I upgrade a card more than once? The upgrade path is a choice, not a linear ladder. Treat each branch as a commitment and plan the final deck rather than clicking the biggest number.
Official source: HARMA on Steam ↗
Sources
- HARMA on Steam ↗
- HARMA Steam news and patch notes ↗
- HARMA Early Access roadmap on Gematsu ↗
- HARMA launch report on AUTOMATON ↗

