RotomArenaAI gaming benchmark • 2 benchmarks

Matches · 2026-10-10 · rules version c3

GPT-6 Luna Decisions vs Perplexity Decider V1.1 27B

Player 1
GPT-6 Luna Decisions
GPT-6 Luna Decisions's team
autopilot 0%
Player 2 · winner
Perplexity Decider V1.1 27B
Perplexity Decider V1.1 27B's team
autopilot 0%

Perplexity Decider V1.1 27B won in 11 turns · 8 s wall clock · 40,220 tokens · $0.002

The replay player loads from play.pokemonshowdown.com. The full battle log is below.

Decisions

  1. Team preview
    GPT-6 Luna Decisions brought Rotomwash,garchomp,gholdengo4,376 tok · 1.2 s · $0.0004

    Lead: 52% BRING: rotomwash, 30% BRING: garchomp, 10% BRING: ninetalesalola Pick 2: 43% BRING: garchomp, 21% BRING: gholdengo, 17% BRING: scizor Pick 3: 90% BRING: gholdengo, 5% BRING: incineroar, 4% BRING: scizor

    Perplexity Decider V1.1 27B brought Annihilape,tinkaton,chandelure4,972 tok · 1.5 s · $0.0001

    Lead: 38% BRING: annihilape, 29% BRING: tinkaton, 15% BRING: palafin Pick 2: 43% BRING: tinkaton, 22% BRING: chandelure, 20% BRING: palafin Pick 3: 47% BRING: chandelure, 27% BRING: palafin, 16% BRING: incineroar

  2. Turn 1
    GPT-6 Luna Decisions used Willowisp1,038 tok · 0.4 s · $0.0001

    44% MOVE: willowisp, 40% MOVE: voltswitch, 10% MOVE: hydropump

    Perplexity Decider V1.1 27B used Closecombat1,163 tok · 0.5 s · $0.0000

    63% MOVE: closecombat, 16% MOVE: earthquake, 12% MOVE: phantomforce

  3. Turn 2
    GPT-6 Luna Decisions switched to Gholdengo1,149 tok · 0.4 s · $0.0001

    21% SWITCH: gholdengo, 19% MOVE: hydropump, 18% MOVE: voltswitch

    Perplexity Decider V1.1 27B used Closecombat1,121 tok · 0.5 s · $0.0000

    96% MOVE: closecombat, 2% SWITCH: tinkaton, 2% SWITCH: chandelure

  4. Turn 3
    GPT-6 Luna Decisions used Shadowball1,185 tok · 0.4 s · $0.0001

    99% MOVE: shadowball, 1% MOVE: makeitrain, 0% MOVE: nastyplot

    Perplexity Decider V1.1 27B switched to Tinkaton1,223 tok · 0.5 s · $0.0000

    72% SWITCH: tinkaton, 25% SWITCH: chandelure, 2% MOVE: closecombat

  5. Turn 4
    GPT-6 Luna Decisions used Makeitrain1,265 tok · 0.4 s · $0.0001

    51% MOVE: makeitrain, 48% MOVE: shadowball, 1% SWITCH: garchomp

    Perplexity Decider V1.1 27B used Gigatonhammer1,399 tok · 0.5 s · $0.0000

    54% MOVE: gigatonhammer, 13% SWITCH: annihilape, 12% SWITCH: chandelure

    Perplexity Decider V1.1 27B switched to Annihilape1,333 tok · 0.5 s · $0.0000

    77% SWITCH: annihilape, 23% SWITCH: chandelure

  6. Turn 5
    GPT-6 Luna Decisions switched to Garchomp1,308 tok · 0.4 s · $0.0001

    50% SWITCH: garchomp, 31% MOVE: shadowball, 6% MOVE: makeitrain

    Perplexity Decider V1.1 27B used Earthquake1,424 tok · 0.5 s · $0.0000

    50% MOVE: earthquake, 28% MOVE: icepunch, 12% MOVE: phantomforce

  7. Turn 6
    GPT-6 Luna Decisions used Earthquake1,309 tok · 0.4 s · $0.0001

    87% MOVE: earthquake, 4% MOVE: dragonclaw, 3% MOVE: stoneedge

    Perplexity Decider V1.1 27B used Earthquake1,321 tok · 0.4 s · $0.0000

    86% MOVE: earthquake, 14% SWITCH: chandelure

  8. Turn 7
    GPT-6 Luna Decisions used Earthquake1,305 tok · 0.4 s · $0.0001

    99% MOVE: earthquake, 1% SWITCH: gholdengo, 0% MOVE: stoneedge

    Perplexity Decider V1.1 27B used Earthquake1,323 tok · 0.6 s · $0.0000

    91% MOVE: earthquake, 9% SWITCH: chandelure

    GPT-6 Luna Decisions switched to Rotom1,160 tok · 0.4 s · $0.0001

    83% SWITCH: rotomwash, 17% SWITCH: gholdengo

  9. Turn 8
    GPT-6 Luna Decisions used Hydropump1,212 tok · 0.4 s · $0.0001

    77% MOVE: hydropump, 11% MOVE: voltswitch, 5% MOVE: willowisp

    Perplexity Decider V1.1 27B switched to Chandelure1,258 tok · 0.5 s · $0.0000

    78% SWITCH: chandelure, 22% MOVE: earthquake

  10. Turn 9
    GPT-6 Luna Decisions used Hydropump1,271 tok · 0.4 s · $0.0001

    78% MOVE: hydropump, 20% MOVE: voltswitch, 1% MOVE: willowisp

    Perplexity Decider V1.1 27B used Fireblastmega1,620 tok · 0.5 s · $0.0000

    59% MOVE: fireblast [MEGA], 13% MOVE: energyball [MEGA], 10% MOVE: shadowball [MEGA]

  11. Turn 10
    GPT-6 Luna Decisions used Hydropump1,294 tok · 0.4 s · $0.0001

    57% MOVE: hydropump, 22% SWITCH: gholdengo, 14% MOVE: voltswitch

    Perplexity Decider V1.1 27B used Shadowball1,457 tok · 0.5 s · $0.0000

    46% MOVE: shadowball, 29% MOVE: fireblast, 22% MOVE: energyball

    GPT-6 Luna Decisions switched to Gholdengo

    100% SWITCH: gholdengo

  12. Turn 11
    Perplexity Decider V1.1 27B used Fireblast1,457 tok · 0.5 s · $0.0000

    53% MOVE: fireblast, 19% MOVE: shadowball, 19% MOVE: energyball

    GPT-6 Luna Decisions used Shadowball1,277 tok · 0.5 s · $0.0001

    93% MOVE: shadowball, 4% MOVE: nastyplot, 2% MOVE: makeitrain