Key takeaways
- Scaling beats burst past the level cap. A race or grimoire that adds a percentage multiplier keeps gaining value as your base stats grow; one that adds flat damage stops mattering once gear outpaces it.
- Utility is worth more than its score suggests in solo play. Sustain and crowd control reduce the number of times you die, and every death costs more time than a few points of damage ever returns.
The best overall pick in Black Grimoire Odyssey is the Dragonkin race paired with the Void grimoire: Dragonkin has the strongest late-game scaling curve of any race, and Void’s defense-shred and execute passives multiply that scaling instead of competing with it. If you are early in the game or playing free-to-play, the Adaptive Human race with the Pyre grimoire clears content faster per hour played and costs far less to reroll. Everything below is ranked on damage, utility and late-game scaling using a 10-point framework, with reroll priority noted for each entry.
How these tiers are scored
The scores below are this article’s own 10-point framework across the three axes that decide endgame value — raw damage output, utility (mobility, crowd control, sustain, team value) and late-game scaling (how much the entry gains per level, per gear tier and per patch). They are not tooltip numbers, and balance patches routinely shift the underlying values, so treat the tiers as role definitions rather than fixed stat lines.
Two rules keep the ranking honest:
- Scaling beats burst past the level cap. A race or grimoire that adds a percentage multiplier keeps gaining value as your base stats grow; one that adds flat damage stops mattering once gear outpaces it.
- Utility is worth more than its score suggests in solo play. Sustain and crowd control reduce the number of times you die, and every death costs more time than a few points of damage ever returns.
Quick comparison: the five picks worth building around
| Pick | Best for | Key scores (DMG / UTIL / LATE) | Reroll priority |
|---|---|---|---|
| Dragonkin + Void | Endgame min-maxing, boss DPS | 8 / 6 / 10 race · 9 / 8 / 10 grimoire | Highest — reroll everything else toward this |
| Revenant + Frost | Free-to-play, solo survival | 7 / 8 / 9 race · 7 / 9 / 8 grimoire | Low — works with no premium rerolls |
| High Elf + Pyre | Burst windows, group raids | 9 / 4 / 7 race · 10 / 5 / 7 grimoire | Medium — strong early, needs utility support later |
| Stormkin + Chronos | PvP, cooldown-heavy builds | 7 / 7 / 8 race · 6 / 10 / 9 grimoire | Medium — best PvP pairing |
| Adaptive Human + Pyre | New accounts, first 10 hours | 6 / 6 / 5 race · 10 / 5 / 7 grimoire | Skip — level up first, reroll later |
Race tier list: every archetype ranked
| Tier | Race archetype | Damage | Utility | Late-game scaling | Where it wins |
|---|---|---|---|---|---|
| S | Dragonkin (draconic, high base HP + elemental breath) | 8 | 6 | 10 | Long fights, boss phases, any build that scales off max HP |
| S | Revenant (undead, self-revive + damage-over-time) | 7 | 8 | 9 | Solo farming, no-healer groups, death-heavy content |
| A | High Elf (caster, highest base magic multiplier) | 9 | 4 | 7 | Short burst windows, raid DPS checks |
| A | Stormkin (mobility, attack-speed and dash passives) | 7 | 7 | 8 | PvP, dodge-timing builds, cooldown cycling |
| A | Beastkin (crit rate and movement speed) | 7 | 5 | 7 | Farming speed, open-world traversal |
| B | Oni (burst melee, high early damage) | 9 | 3 | 6 | Early and mid game only — flat bonuses fall off |
| B | Forgekin (defense, crafting and inventory utility) | 5 | 9 | 6 | Support roles, resource farming, group play |
| B | Adaptive Human (balanced, cheap to reroll) | 6 | 6 | 5 | Learning the game, no-commitment accounts |
The trade-off is consistent: the S-tier races start slower. Dragonkin’s breath passives only become dominant once your max-HP pool is large enough for percentage-based damage to matter, which usually means after the mid-game gear tier. Oni and High Elf feel far stronger for the first several hours and then plateau, which is why so many players reroll too early in one direction and too late in the other.
Grimoire tier list: ranked by damage, utility and scaling
| Tier | Grimoire | Damage | Utility | Late-game scaling | Notes |
|---|---|---|---|---|---|
| S | Void | 9 | 8 | 10 | Defense shred and execute; the best multiplier in the game |
| S | Chronos | 6 | 10 | 9 | Cooldown reduction compounds with every other ability |
| A | Pyre | 10 | 5 | 7 | Highest raw numbers; weakest utility |
| A | Frost | 7 | 9 | 8 | Control and slow; safest solo pick |
| A | Storm | 8 | 6 | 8 | Chain damage; excels in multi-target farming |
| B | Blood | 8 | 4 | 7 | Lifesteal sustain, but health-cost mechanics punish mistakes |
| B | Verdant | 6 | 9 | 6 | Summons and regeneration; summons scale poorly late |
| B | Aegis | 4 | 9 | 6 | Shields and damage reduction; a group pick, not a solo one |
| C | Arcane (starter) | 5 | 5 | 4 | Fine for the tutorial, replace by the mid game |
Void and Chronos sit above the raw-damage grimoires because their effects are multiplicative — a defense shred increases everyone’s damage, and cooldown reduction increases how often every ability fires. Pyre wins any single-target damage comparison in a vacuum, but it brings nothing to a fight that a teammate is not already providing.
Best race and grimoire combos by situation
| Your situation | Race | Grimoire | Why this pairing |
|---|---|---|---|
| Brand-new account, first 10 hours | Adaptive Human | Pyre | Cheapest reroll cost, fastest clear speed while learning systems |
| Free-to-play, limited rerolls | Revenant | Frost | Self-revive plus control covers mistakes without premium items |
| Group raids and boss DPS checks | High Elf | Pyre | Maximum damage inside short vulnerability windows |
| Endgame min-maxing | Dragonkin | Void | Two multiplicative scaling sources stacking on the same build |
| PvP and duels | Stormkin | Chronos | Mobility plus cooldown cycling beats raw damage against players |
| Solo open-world farming | Beastkin | Storm | Movement speed and chain damage clear packs fastest |
Reroll advice: when to reroll, when to stop
Rerolling is where most accounts lose progress. The math is straightforward if you state the assumptions: suppose a single race reroll costs 150 reroll tokens and an S-tier outcome lands roughly 1 time in 40. The chance of hitting at least one S-tier result across repeated attempts is 1 − (39/40)n.
| Rerolls attempted | Tokens spent | Chance of at least one S-tier |
|---|---|---|
| 10 | 1,500 | 22% |
| 20 | 3,000 | 40% |
| 28 | 4,200 | 50% |
| 60 | 9,000 | 78% |
| 91 | 13,650 | 90% |
If your dailies and farming yield roughly 250 tokens a day, a coin-flip chance at an S-tier race costs about 17 days of income, and a 90% confidence level costs about 55 days. That is the number to weigh against simply playing an A-tier race well.
- Reroll your race first if it is below B tier. A C-tier race costs you on every single encounter.
- Stop rerolling once you hold an A-tier. In this framework, moving from A to S is worth roughly 6–9% endgame damage, while moving from a B-tier grimoire to an S-tier one is worth 12–18% because grimoire passives multiply more of your kit.
- Never reroll before the level cap. Unlocking the remaining ability slots changes which stats matter, and you can end up discarding a roll that was actually correct.
What falls off first in the late game
Flat damage bonuses are the first thing to become irrelevant — Oni’s early burst and Arcane’s starter scaling both stop keeping pace once gear multipliers arrive. Critical chance is the second, because it hits a practical ceiling and every point beyond it is wasted. Percentage-based effects — max-HP scaling, defense shred, cooldown reduction — are the last to fall off and the ones worth building around.
Ownership realities matter here too. Respec scrolls and stat-reset items are the real consumables of this game; a single reroll decision can cost weeks of accumulated tokens, and there is no way to reverse it cheaply. The most common mistakes are rerolling two things at once (you lose track of which change actually helped), chasing damage while ignoring sustain in solo content, and treating a tier list as permanent when a balance patch can move an entire archetype by a full tier.
Frequently Asked Questions
What is the best race in Black Grimoire Odyssey?
Dragonkin is the best race for endgame play because its percentage-based passives scale with your max-HP pool and keep gaining value after the level cap. Revenant is the better pick for free-to-play and solo players, since its self-revive removes the cost of deaths without requiring any premium currency. Adaptive Human remains the right choice for a brand-new account.
Which grimoire has the highest damage?
Pyre has the highest raw single-target damage numbers, but Void generally produces more total damage in real fights because its defense shred increases the damage of every other ability in your kit. In multi-target farming, Storm often outperforms both because chain effects hit several enemies at once.
Should I reroll my race or my grimoire first?
Reroll the race first if it is below B tier, because a weak race affects every encounter you play. Once you hold an A-tier race, switch attention to the grimoire: grimoire upgrades in this framework are worth roughly twice the endgame damage gain of an A-to-S race upgrade.
Is Adaptive Human good in the late game?
No — Adaptive Human is a starting race, not an endgame one. Its balanced stat spread and cheap reroll cost make it ideal for learning the systems, but its scaling score is the lowest on the list and it loses ground to Dragonkin and Revenant once percentage-based multipliers take over.
How many rerolls does it take to get an S-tier race?
With a 1-in-40 S-tier rate, 28 rerolls give you roughly a 50% chance and 91 rerolls give you roughly a 90% chance. Budget for the outcome you actually need rather than rerolling until you feel lucky, because the token cost of the last 10% of certainty is usually better spent on gear.
Do race and grimoire bonuses stack additively or multiplicatively?
Damage bonuses from race and grimoire generally combine multiplicatively, which is why pairing two scaling sources — such as Dragonkin with Void — produces more total output than pairing a scaling source with a flat-damage one. Always check the current in-game stat panel after a balance patch, since the game’s official patch notes are the only reliable source for the current numeric caps.
