The Daily Run is live in Early Access: what it is and how to unlock it
You open the game on a Tuesday morning and the menu offers a run you did not design: three modifiers stacked on top of each other, one shared seed, one Ascension level, and a leaderboard with your friends' names on it. That is the Daily Run in Slay the Spire 2, and the first thing worth saying plainly is that it exists. It shipped with the Early Access systems rather than arriving later, so if you played the first game during its first year and remember waiting for the Daily Climb, the situation is different this time. As of v0.111 beta (September 2026), both the Daily Run and a fully featured Custom Mode are in the game.
Getting in is the confusing part, because the two modes unlock separately and the documentation disagrees. The wiki states that Custom Mode unlocks after you defeat Glory, the Act 3, three times. The Daily Run's own in-game text, quoted by players on the Steam forums, asks you to play a single run with all characters and beat Act 3 to unlock the daily run. A Reddit thread on the unlock reports that order seems to matter: players who beat Act 3 before trying every character did not get the unlock until they cleared it again afterward. Treat that order detail as community-reported, but if your daily button is greyed out and you have a Glory clear, the checklist is: all five characters played at least once, then an Act 3 win.
Note the asymmetry between the two conditions. Three Glory clears is a committed player's milestone, roughly your first week or two with the game. The daily condition, one clear per the in-game text plus roster coverage, is lighter, and the character chain means playing through Ironclad, Silent, Regent and Necrobinder at least once anyway, so most players will naturally satisfy the coverage half while working toward the Defect. Neither unlock mode grants achievements or Timeline progress, which tells you what they are for: not advancement, but repetition and practice.
The frame for using them well is different from ranked play. A normal run is an investment; you route carefully, you protect the timeline score, you climb Ascension. A daily or custom run is a laboratory. The modifiers are the instruments. The rest of this guide covers how the daily is structured, every modifier in the current pool with training uses for each, how seeds work after the v0.107.1 random number rework, and a practice plan that turns twenty minutes a day into measurable improvement at any Ascension level you are stuck on.
Three modifiers, one seed, one Ascension: the shape of a daily
Every daily run has the same skeleton. Three modifiers are chosen for the day, and every player in the world gets the same three, the same seed and the same Ascension level. The wiki hedges on whether the selection is fully random or curated, and Mega Crit has not clarified it, so expect occasional days where the three clearly interact (a Draft day with Cursed Run, for example) and days where they have nothing to do with each other. The seed is the real equalizer: your map, your card rewards, your relic rolls and your shop stock all match everyone else's, so leaderboard gaps are decision gaps, not luck gaps.
Scoring was deliberately simplified in v0.103.0 (April 2026) and the order matters: first whether you won, then how many badges you accrued during the run, then completion speed. That ordering rewards clean, fast, badge-heavy wins and makes slow grinding strictly worse, which changes how you should play a daily relative to a normal run. Badges, the in-run achievement system introduced across v0.102 and Major Update 1, are the tiebreaker currency, so plays that earn badges (big single turns, fast kills, debuff coverage) have extra value on the ladder even when they cost a little efficiency.
The leaderboard is friends-only, and it has been since that same v0.103.0 patch. There is no global board yet; the developers have said a global view is planned for later and, notably, that they do not intend to build invasive anti-cheat to police it, which is the honest reason the friends-first model exists. Beta branch and main branch also keep separate leaderboards, so if you play the beta for v0.111 content, your daily history does not merge with your main-branch friends' list. Check which branch you are on before wondering why a friend's score is missing.
Dailies are also playable in multiplayer lobbies, which is one of the quiet best features of the mode: you and a friend get the same modifiers and seed and fight the same run together, with co-op enemy scaling applied on top. The v0.111 patch fixed a lingering bug where players joining a loaded daily lobby showed as unready, so the flow is stable as of this writing.
One practical warning: because everyone shares the seed, daily discussion online is spoiler-heavy by afternoon in most timezones. If you want the puzzle clean, play the daily before you read Reddit. If you want to learn, play it blind, then compare your route and picks against a friend's replay of the same seed. The delta between two runs of one seed is the most instructive artifact this game produces.
Custom Mode: the sandbox behind three Glory clears
Custom Mode is the laboratory proper, and it unlocks at three total Act 3 clears, which the wiki and the Timeline system both track. Once open, it lets you enable any number of modifiers at once, set the run's seed manually and set the Ascension level to anything you have unlocked. A Randomize button was added in v0.108 for players who want the chaos without the menu work. Nothing you do in Custom Mode or the Daily Run counts toward achievements or the Timeline's Epoch progression, a rule in place since roughly launch and reaffirmed when the leaderboards were rebuilt, so treat every custom run as consequence-free.
The seed control is the feature that changes how you practice. In a normal run, you cannot distinguish a bad decision from a bad roll, because the roll is never repeated. In Custom Mode you can fix the seed, play the run, then start the identical run again and take the other line at the branch point: the fight you skipped, the card you passed, the elite you feared. Same map, same rewards, same shop stock, different choice. Two runs like that teach more than twenty fresh ones, because the feedback is controlled.
The modifier menu is also a difficulty scaler that is finer-grained than Ascension. If A9 kills you with raw damage but A8 bores you, a custom run at A8 with Big Game Hunter (elite swarm) or Murderous (both sides start with Strength) sits between them. If your problem is economic, Midas doubles enemy gold payouts while banning Smith upgrades, which forces spending discipline without changing fight math. You can build a custom difficulty that targets your specific weakness instead of accepting the ladder's fixed steps.
Two structural notes. First, several of the strongest modifiers disable Neow's starting bonus, and the game marks which ones; deck-replacement modifiers like Draft and Insanity all do, so runs that already rewrite your deck skip the Neow layer too. Second, the class-pool modifiers can be stacked, which is the only legitimate way to build a five-color deck outside of Prismatic Gem shenanigans, and combining them with Draft produces something close to a cube-draft format inside a roguelike.
Finally, a housekeeping habit worth forming: name your practice runs. Keep a note of the seed, the modifier set, the question you were testing and the answer you got. The mode makes experimentation free, but only writing makes it cumulative.
Draft, Sealed Deck, Specialized, Insanity and All Star: replacing the starting deck
Five modifiers replace your starting deck outright, and they are the closest thing STS2 has to limited-format card games. All five disable Neow's starting bonus, and three of them (Draft, Sealed Deck, Insanity) also stop Darv from offering Pandora's Box, since transforming all Strikes and Defends is meaningless when you have none.
Draft gives you ten card reward picks to build a deck from scratch, one at a time from generated offers. It is a pure pick-discipline trainer: with no starter baggage, every card is a real choice, and deck curve, upgrade targets and removal needs all have to be planned ten picks ahead. If your normal-run problem is taking cards you do not need, a week of Draft runs will sand that habit off, because the cost of a greedy pick is visible within two fights.
Sealed Deck is Draft's evaluation cousin: you see 30 cards and choose 10. The skill is reading a whole pool at once and finding the deck hiding inside it, rather than reacting to offers sequentially. Sealed also teaches rarity economics, since a sealed pool contains the same rarity spread as real rewards, and passing the rare bomb for the third copy of a working common is a lesson every ladder player eventually needs.
Specialized starts you with five copies of a single card. It is the degenerate stress test: five copies of a block card, five copies of a scaling attack, whatever the game deals. Decks like this expose how much of a normal run is carried by card quality versus card synergy, and they are genuinely useful for learning a single card's ceiling, including its upgraded form, because you will play it a dozen times per fight.
Insanity deals you a random 30-card deck, which is the variance trainer. You cannot curve it, you cannot thin it cheaply, and you have to win with what arrives. Insanity runs teach improvisation, energy discipline with awkward hands, and honest evaluation of cards you never pick voluntarily. Players who claim a card is unplayable usually have not been forced to build around it five times in a row.
All Star starts you with five Colorless cards, the pool most players ignore entirely. Colorless cards in STS2 include tools, situational answers and the enchantment-adjacent oddities, and All Star is the format that finally teaches you their exchange rates. As a bonus, colorless familiarity transfers directly to normal runs, where a well-timed colorless pick off a shop or event solves problems your class pool cannot.
Flight, Vintage, Deadly Events, Midas, Night Terrors, Terminal: rewriting the map
The second bucket of modifiers attacks the map layer rather than the deck, and each one trains a different routing muscle.
Flight removes path restrictions entirely: you may ignore connections when choosing the next room. This is the boss-practice modifier. Fix a seed, enable Flight, and you can walk a direct line to whichever act boss kills you, fight it, and walk the same line again with a different pick pattern. It also exposes how much value normal runs hide in path freedom, since most players under-use the flexible middle rows of the map once an act plan is set.
Vintage makes normal enemies drop relics instead of cards. It is a relic-literacy course: by Act 2 you have more relics than a normal run sees in three wins, and you learn their actual exchange rates by living inside them. The cost is card progression, so Vintage runs also train winning with under-upgraded decks, which is its own useful skill at high Ascension.
Deadly Events lets Unknown rooms contain elites while raising the odds of treasure rooms. It is the risk-assessment modifier. Normal STS2 pathing already weights the value of a question mark against known node types; with elites in the pool, every question mark becomes a real gamble, and treasure weighting makes some of those gambles profitable. If your normal-run weakness is over-pathing into the unknown, this is the corrective.
Midas doubles enemy gold drops to 200 percent and bans Smith at rest sites. The gold flood teaches spending discipline: you can afford nearly everything, so what you actually buy reveals your priorities. The Smith ban is the sharp edge, because it removes the default campfire option and forces creative use of rest, dig, and whatever your relics unlock. Midas at A6-plus is also the fastest legal way to study the Inflation economy, since you finally have the budget to feel the removal pricing curve.
Night Terrors makes resting heal all HP at the cost of 5 Max HP per rest. It converts the resource question from short-term HP to long-term maximums, which is exactly the trade high Ascension runs already force implicitly. Terminal charges 1 Max HP per room entered and grants 5 Plating at combat start; since v0.104 it can no longer drop you below 1 HP, so it is survivable by design. Terminal teaches route compression: every extra room has a literal price, and finishing an act two rooms sooner is worth 2 Max HP forever.
Hoarder, Cursed Run, Big Game Hunter and Murderous: pressure-testing the engine
The third bucket modifies combat and card acquisition, and these are the modifiers that stress-test a deck's engine rather than its picks.
Hoarder adds two copies of every card you add to your deck and disables card removal at the Merchant entirely. Every card pick becomes three cards, so drafting discipline stops being optional: one careless pick is now three dead draws you cannot pay to remove. Hoarder is the deck-thinning teacher in reverse, the format where bloat is guaranteed and the only lever is taking fewer cards in the first place. Exhaust strategies, which delete cards mid-combat, become proportionally stronger, and any archetype that wants copies (scaling attacks, consistent block) gets free redundancy.
Cursed Run adds a random curse at the start of each act, drawing from a ten-curse pool: Clumsy, Debt, Decay, Doubt, Guilty, Injury, Normality, Regret, Shame and Writhe. Three curses across a run is a real tax, and the pool's variety means you cannot pre-plan around a specific one. The training value is generic curse math: which curses are deck damage (Injury, Decay), which are tempo damage (Clumsy, Regret), and which rewrite a turn (Normality). STS2's Afflictions already put debuffs on cards you own, so curse literacy transfers directly.
Big Game Hunter makes elites swarm the map and drop better rewards. Where Ascension 1 spreads roughly 60 percent more elites across the whole run passively, Big Game Hunter goes much further and pays you for it. This is the elite-mastery format: elite pools per act, their intent patterns, and which of your decks can chain three elites without collapsing. Relic income explodes, which also makes it the fastest format for studying relic synergies in combination.
Murderous gives you and every enemy 3 Strength at the start of each combat. It is a symmetric tempo shift that changes every damage breakpoint in the game: hallway fights get faster in both directions, block math shifts, and fights that were slow grinds become races. Murderous is the modifier for practicing aggressive openings, since the side that uses the free Strength first usually wins.
Stacking these is where custom runs get interesting: Big Game Hunter plus Murderous is an elite-gauntlet race format; Hoarder plus Cursed Run is a maximal-bloat discipline test. The Randomize button from v0.108 will happily hand you worse combinations than either, and those runs still teach something, mostly about knowing when to abandon a plan.
Class card modifiers: five ways to break the color pie
Five separate modifiers, one per class, add that class's cards to your reward pools: Ironclad Cards, Silent Cards, Regent Cards, Necrobinder Cards and Defect Cards. The wiki notes the borrowed cards appear in combat rewards and shop stock but not from potions or other card-generating effects, so the flood is filtered. Enable one on top of your normal character and you are effectively playing a two-color deck, which is the closest thing the sequel has to the first game's Prismatic Shard shenanigans, and there is an in-run analog too: the Prismatic Gem Ancient boon from Orobas achieves a similar effect inside normal runs.
The strategic uses go beyond novelty. Each class contributes what your base class lacks: the Silent's cards bring poison and discard utility to anyone; the Necrobinder brings Doom, which executes enemies whose Doom exceeds their HP at end of turn and notably bypasses Block, Slippery and Intangible, a hard answer to several boss problems; the Ironclad brings Exhaust synergies and Strength; the Defect brings orbs and focus-free scaling; the Regent brings Stars, Forge and Vigor lines that no other class can fuel efficiently.
That last point is the caveat. Some class mechanics are parasitic: Stars fuel Regent cards, Osty is required for Necrobinder attack cards to resolve, orbs need evoke and channel density. Cross-class drafting works best when you take self-contained packages (a poison stack plus its support, a Strength line with its payoff attacks) rather than scattered singles from a foreign engine. The wiki documents this boundary directly: cards like Osty attacks resolve with no effect for anyone without Osty summoned, and colorless multi-class pickups through Prismatic Gem or the Colorful Philosophers event hit the same walls.
For practice purposes, the class modifiers are archetype school. Stuck on a matchup with your main? Build the Silenced-Ironclad hybrid in Custom Mode and learn what poison timing actually looks like from the other side. Curious whether Doom solves your Aeonglass problem before committing ladder attempts to a Necrobinder climb? Draft a Defect run with Necrobinder Cards enabled and find out in twenty minutes.
One interaction worth knowing: class-pool modifiers stack with Draft and Sealed Deck, so the deepest custom formats, a five-color draft, are two checkboxes away. Those runs are chaotic, occasionally brilliant, and completely progression-free, which is precisely the point.
Seeds, the v0.107.1 RNG rework, and what the community got wrong first
Custom runs start from a random seed or one you type in, and the seed drives the map layout, card and relic rolls and shop generation. That is the mechanics. The interesting part of the story is what the community discovered about the seeds themselves.
Through the first months of Early Access, players noticed correlations between random streams: community member tckmn demonstrated that the Act 1 generation roll could be used to predict the curse attached to Neow's Bones, meaning separate systems were quietly reading from correlated number streams. Mega Crit reworked the random number generation to xoshiro256 in Major Update 2 (v0.107.1, June 2026), alongside Steam Workshop support, which dissolved those correlations. Two practical consequences: seed knowledge from before late June 2026 is stale and community-documented seed exploits from the early months no longer work, and current seeds are stable, so a seed shared today produces the same run for everyone on the same branch.
Branch is the fine print. Beta and main versions have diverged repeatedly through v0.108 to v0.111, and any balance change to a card or enemy alters what a seed produces downstream of that change, even when the underlying rolls are identical. If you and a friend are comparing seeds, confirm you are on the same branch and version first, or the comparison silently breaks.
The wiki maintains a table of seeds with unique effects, but it is an unfinished placeholder, with most entries reading as unknown. Special named seeds with unique effects, a tradition carried over from the first game, may exist in STS2 as well; nothing verified has surfaced, so treat any seed you see advertised with special powers as unverified until you test it. Community sites that publish curated seeds do exist, and their contents are community-reported rather than vetted.
The training use of seeds is A/B testing, and it deserves the emphasis. Fix a seed. Play it once following your instinct, once following the plan you suspect is better. The second run answers the question the first run asked. Reuse the same seed across a week with different modifiers to isolate one variable at a time. This is the only tool in the game that converts anecdote into evidence, and it is sitting behind a text box in a menu most players open once.
A four-week practice plan built from dailies and custom runs
Here is a concrete curriculum, assuming twenty to forty minutes a day, for turning the two modes into improvement rather than variety.
Week one: pick discipline. Play the daily every day, blind, before reading anything about it. After each run, write one line: the floor where the run was decided and the pick that caused it. Then replay that day's seed in Custom Mode at the same Ascension and take the other line at the decisive branch. Two runs a day, one decision isolated. Also run three Drafts during the week; Draft exposes pick habits faster than any other format because there is no starter deck to hide behind.
Week two: boss reps. You know which boss ends your runs, and in Custom Mode you can practice it on demand: fix a seed, enable Flight, path straight to the boss, and play the fight three times with three different lines. Learn the opening turns cold, because STS2 bosses run fixed, telegraphed loops and the first three turns decide most of the fight. The Queen's Amalgam clock, Test Subject's phase resets, Aeonglass's escalating cycle: these are patterns, and patterns are learnable faster in a lab than across ladder attempts.
Week three: stress tests. Assign one pressure modifier per session. Hoarder for pick discipline, Cursed Run for curse math, Terminal for route compression, Midas for spending, Murderous for tempo. Play each at your current Ascension level, not below it; the point is to hold your standard while the ground shifts. If a modifier reveals a specific hole (your block math collapses under Murderous), spend the next session on that hole directly with a fixed seed.
Week four: competition. Keep playing the daily blind, but start racing it: the scoring system ranks wins first, badges second, speed third, so practice clean fast lines and badge-earning plays. Compare against your friends list, since leaderboards are friends-only, and discuss routes afterward. If you have a regular group, run the daily in a multiplayer lobby once; co-op scaling plus daily modifiers is the hardest casual format in the game.
Measure the month honestly: win rate at your fixed Ascension, average death floor, and potion-slots-left-at-death (a full rack at death means you died with unused resources). Progress in this game is boring and compounding, which is exactly why the laboratory matters: the ladder charges full price for every lesson, and the sandbox gives most of them away.
What daily play progresses (and what it deliberately does not)
The boundary is worth stating cleanly because it surprises people: Daily Runs and Custom Runs grant no achievements and no Timeline Epoch progress. The rule dates to the Early Access launch window and was restated when the leaderboard rework landed in v0.103.2 territory. Your Timeline score, your Epoch unlocks, your character unlocks, your Ascension ladder, all of it moves only in normal runs (and, separately, in multiplayer). The modes are designed as playgrounds, not treadmills.
That design choice is more useful than it first looks. Because nothing is at stake, the daily is the one place you can try the greedy line, the weird potion timing, the archetype you do not trust. Fear of wasting a good run is the main reason players plateau at a comfortable style, and the daily deletes that fear by construction. You cannot waste a daily; there is a new one tomorrow with new constraints.
The social layer is what the daily does progress: your standing against friends. With friends-only leaderboards and the won-then-badges-then-speed ranking, a friends group effectively runs a private competitive league on shared seeds. That is a different skill than ladder climbing: matching a known field on a known puzzle rewards preparation and speed over raw consistency. Global leaderboards are planned, and the developers have been explicit that they will not ship invasive anti-cheat with them, so expect the global scene to stay honor-system casual even when it arrives.
Two practical footnotes for planners. First, daily and custom runs still respect your unlocked content: Ascension levels in Custom Mode are capped at what you have unlocked, and the modifiers themselves are all available once the mode opens, no per-modifier unlocks. Second, co-op daily lobbies are stable as of v0.111 after the unready bug fix, so the mode is a legitimate group activity; just remember co-op has its own Ascension track and the daily runs on whatever the lobby sets.
The summary of the whole system: normal runs build your account, Ascension builds your ceiling, and the daily and custom sandbox builds the player. Twenty minutes of controlled practice beats two hours of ladder grinding for most people, because the feedback loop is tighter and the cost of a bad experiment is zero. Play the daily blind in the morning, replay its seed with a different plan in the evening, and check the leaderboard not to see who won but to see what the winners did differently with the same cards you were dealt.