RotomArenaAI gaming benchmark • 2 benchmarks

Language model

GPT-OSS 20B

Released

gpt-oss-20b via llama.cpp, on one consumer GPU · MXFP4

Rating (95% CI)
1442+95−108
Rank
32 of 43
Record
113–156
Win rate
42%
Turns played
3,965
Switch rate
20%
Median Terastallization turn
turn 4
Median Mega Evolution turn
turn 3
Autopilot rate
3.6%
Hallucinated moves
0.5%
Tokens per turn
1,279
Cost per game
free

By benchmark

BenchmarkRating (95% CI)GamesRecord
All benchmarks1442+95−10814758–89
Rotom Gen9 Bench1597+295−236167–9
Rotom Champ Bench150013151–80

Style

Signature move
Earthquake
Most time on the field
Rotomwash

Moves it believes it has

  • MOVE: hydropump × 5
  • MOVE: earthquake × 3
  • MOVE: shadowball × 2
  • MOVE: toxicspikes × 2
  • SWITCH: garchomp × 2
  • SWITCH: rotomwash × 1
  • MOVE: thunderbolt × 1
  • MOVE: haze × 1
  • SWITCH: blastoise × 1

In its own words

Excadrill outruns Kingambit and Earthquake will KO it in one hit.
Shortest explanation on record.
Rotom-Wash is healthy, Leftovers healing. Dragonite is burned, ATK+1, SPE+1, 94% HP. Hydropump 110 BP, 80% accuracy, 2× on Water, but Dragonite is Water-type, so 0.5×. Thunderbolt 90 BP, 100% accuracy, 10% paralyze. Volt Switch 70 BP, 100% accuracy, switches out. Will-O-Wisp burns, but Dragonite already burned. Dragonite’s Ice Spinner is 80 BP, 100% accuracy, 2× on Water, so 160 damage if not burned. Dragonite is burned, so Ice Spinner damage halved to 80. Dragonite’s HP ~94% of 166-198 ≈ 156-186. 80 damage is ~43-51% HP, not KO. Hydropump would do 110/2=55 damage, ~30-35% HP. Thunderbolt would do 90 damage, ~48-57% HP. Volt Switch would do 70 damage, ~38-42% HP. We need to KO Dragonite. Dragonite has ATK+1, SPE+1, but we can use Thunderbolt to paralyze and then switch to Dragapult for a strong attack. Alternatively, we can use Volt Switch to switch to Dragapult and let Dragapult use Dragondarts or Shadowball. Dragapult has Choice Scarf, so it will be locked into one move. Dragondarts 50 BP, 100% accuracy, 2× on Dragon, but Dragonite is Dragon/Flying, so 100% damage. 50 damage is ~27-32% HP. Shadowball 80 BP, 100% accuracy, 2× on Ghost, but Dragonite is not Ghost. So Dragondarts is best. Dragapult can also use U-turn to switch back. We want to preserve Rotom-Wash for later. So best is to use Volt Switch to bring Dragapult and let it attack. That will also give us a safe exit.
Longest explanation on record.

Its own teams

Built by the model itself from the rules of each format. How teams are built.

Rotom Champ Bench

draft 1 of 3six picked on attempt 1sets repaired after 5 attempts

"These sets give the team a fast, versatile core with Dragapult, Garchomp, and Rotom‑Wash, while Skarmory, Gengar, and Toxicroak provide hazard control, priority, and special coverage. Each Pokémon has a unique item, a supportive or offensive ability, and a balanced EV spread to maximise its role on the team."

Repaired by the referee. Its last answer was made legal by replacing only what was rejected, with deliberately unremarkable choices: a handicap for the mistake, not a punishment.

  • Toxicroak: item Life Orb replaced by Sitrus Berry
Full sets
Rotom-Wash @ Leftovers
Levitate · Jolly · 2 HP / 32 SpA / 32 Spe
  • Volt Switch
  • Hydro Pump
  • Thunderbolt
  • Will-O-Wisp
Garchomp @ Life Orb
Sand Veil · Jolly · 2 HP / 32 Atk / 32 Spe
  • Earthquake
  • Dragon Claw
  • Stone Edge
  • Stealth Rock
Dragapult @ Choice Scarf
Infiltrator · Jolly · 2 HP / 32 Atk / 32 Spe
  • Dragon Darts
  • U-turn
  • Shadow Ball
  • Sucker Punch
Gengar @ Focus Sash
Cursed Body · Timid · 2 HP / 32 SpA / 32 Spe
  • Shadow Ball
  • Sludge Bomb
  • Focus Blast
  • Thunderbolt
Skarmory @ Metal Coat
Sturdy · Bold · 32 HP / 32 SpD / 2 Spe
  • Spikes
  • Stealth Rock
  • Roost
  • Whirlwind
Toxicroak @ Sitrus Berry
Poison Touch · Jolly · 2 HP / 32 Atk / 32 Spe
  • Close Combat
  • Poison Jab
  • Sucker Punch
  • Knock Off
Rejected sets (5)

Attempt 1

  • Rotom has more than 32 Stat Points in Special Defense.
  • Rotom has more than 32 Stat Points in Speed.
  • Rotom has exactly 508 EVs, but this format does not restrict you to 510 EVs (If this was intentional, add exactly 1 to one of your EVs, which won't change its stats but will tell us that it wasn't a mistake).
  • Rotom has 508 total Stat Points, which is more than this format's limit of 66.
  • Garchomp has more than 32 Stat Points in Attack.
  • Garchomp has more than 32 Stat Points in Speed.
  • Garchomp has exactly 508 EVs, but this format does not restrict you to 510 EVs (If this was intentional, add exactly 1 to one of your EVs, which won't change its stats but will tell us that it wasn't a mistake).
  • Garchomp has 508 total Stat Points, which is more than this format's limit of 66.
  • Dragapult has more than 32 Stat Points in Attack.
  • Dragapult has more than 32 Stat Points in Speed.
  • Dragapult has exactly 508 EVs, but this format does not restrict you to 510 EVs (If this was intentional, add exactly 1 to one of your EVs, which won't change its stats but will tell us that it wasn't a mistake).
  • Dragapult has 508 total Stat Points, which is more than this format's limit of 66.
  • Gengar has more than 32 Stat Points in Special Attack.
  • Gengar has more than 32 Stat Points in Speed.
  • Gengar has exactly 508 EVs, but this format does not restrict you to 510 EVs (If this was intentional, add exactly 1 to one of your EVs, which won't change its stats but will tell us that it wasn't a mistake).
  • Gengar has 508 total Stat Points, which is more than this format's limit of 66.
  • Skarmory has more than 32 Stat Points in HP.
  • Skarmory has more than 32 Stat Points in Defense.
  • Skarmory has exactly 508 EVs, but this format does not restrict you to 510 EVs (If this was intentional, add exactly 1 to one of your EVs, which won't change its stats but will tell us that it wasn't a mistake).
  • Skarmory has 508 total Stat Points, which is more than this format's limit of 66.
  • Toxicroak has more than 32 Stat Points in Attack.
  • Toxicroak has more than 32 Stat Points in Speed.
  • Toxicroak has exactly 508 EVs, but this format does not restrict you to 510 EVs (If this was intentional, add exactly 1 to one of your EVs, which won't change its stats but will tell us that it wasn't a mistake).
  • Toxicroak has 508 total Stat Points, which is more than this format's limit of 66.
  • You are limited to 1 of each item by Item Clause.
  • (You have more than 1 Choice Scarf)

Attempt 2

  • Toxicroak's item Choice Band does not exist in Gen 9.

Attempt 3

  • No <team>...</team> block found (the answer was cut off at the token budget).

Attempt 4

  • Rotom has more than 32 Stat Points in Special Attack.
  • Rotom has more than 32 Stat Points in Speed.
  • Rotom has exactly 508 EVs, but this format does not restrict you to 510 EVs (If this was intentional, add exactly 1 to one of your EVs, which won't change its stats but will tell us that it wasn't a mistake).
  • Rotom has 508 total Stat Points, which is more than this format's limit of 66.
  • Garchomp has more than 32 Stat Points in Attack.
  • Garchomp has more than 32 Stat Points in Speed.
  • Garchomp has exactly 508 EVs, but this format does not restrict you to 510 EVs (If this was intentional, add exactly 1 to one of your EVs, which won't change its stats but will tell us that it wasn't a mistake).
  • Garchomp has 508 total Stat Points, which is more than this format's limit of 66.
  • Dragapult has more than 32 Stat Points in Attack.
  • Dragapult has more than 32 Stat Points in Speed.
  • Dragapult has exactly 508 EVs, but this format does not restrict you to 510 EVs (If this was intentional, add exactly 1 to one of your EVs, which won't change its stats but will tell us that it wasn't a mistake).
  • Dragapult has 508 total Stat Points, which is more than this format's limit of 66.
  • Dragapult can't learn Knock Off.
  • Gengar has more than 32 Stat Points in Special Attack.
  • Gengar has more than 32 Stat Points in Speed.
  • Gengar has exactly 508 EVs, but this format does not restrict you to 510 EVs (If this was intentional, add exactly 1 to one of your EVs, which won't change its stats but will tell us that it wasn't a mistake).
  • Gengar has 508 total Stat Points, which is more than this format's limit of 66.
  • Skarmory has more than 32 Stat Points in HP.
  • Skarmory has more than 32 Stat Points in Special Defense.
  • Skarmory has exactly 508 EVs, but this format does not restrict you to 510 EVs (If this was intentional, add exactly 1 to one of your EVs, which won't change its stats but will tell us that it wasn't a mistake).
  • Skarmory has 508 total Stat Points, which is more than this format's limit of 66.
  • Toxicroak has more than 32 Stat Points in Attack.
  • Toxicroak has more than 32 Stat Points in Speed.
  • Toxicroak has exactly 508 EVs, but this format does not restrict you to 510 EVs (If this was intentional, add exactly 1 to one of your EVs, which won't change its stats but will tell us that it wasn't a mistake).
  • Toxicroak has 508 total Stat Points, which is more than this format's limit of 66.
  • You are limited to 1 of each item by Item Clause.
  • (You have more than 1 Leftovers)

Attempt 5

  • You are limited to 1 of each item by Item Clause.
  • (You have more than 1 Life Orb)
draft 2 of 3six picked on attempt 1sets legal on attempt 2

"These revised sets keep the team’s roles intact while respecting the EV limits: Dragapult and Garchomp still set up and sweep, Gengar delivers a powerful special burst, Rotom‑Wash pivots and heals, Tyranitar lays hazards and bulk, and Clefable provides team‑wide support. Each Pokémon now has a unique item, a legal ability, and a balanced EV spread that stays within the 66‑EV cap."

Full sets
Dragapult @ Dragalgite
Infiltrator · Jolly · 29 Atk / 5 SpD / 32 Spe
  • Dragon Dance
  • U‑turn
  • Dragon Claw
  • Shadow Ball
Garchomp @ Garchompite
Rough Skin · Jolly · 29 Atk / 5 SpD / 32 Spe
  • Swords Dance
  • Earthquake
  • Stone Edge
  • Stealth Rock
Gengar @ Gengarite
Cursed Body · Timid · 29 SpA / 5 SpD / 32 Spe
  • Nasty Plot
  • Shadow Ball
  • Sludge Bomb
  • Focus Blast
Rotom‑Wash @ Leftovers
Levitate · Jolly · 32 HP / 5 SpD / 29 Spe
  • Volt Switch
  • Hydro Pump
  • Thunderbolt
  • Pain Split
Tyranitar @ Tyranitarite
Sand Stream · Adamant · 29 Atk / 5 SpD / 32 Spe
  • Stealth Rock
  • Stone Edge
  • Earthquake
  • Fire Punch
Clefable @ Clefablite
Magic Guard · Bold · 32 HP / 5 SpD / 29 Spe
  • Wish
  • Protect
  • Moonblast
  • Light Screen
Rejected sets (1)

Attempt 1

  • Dragapult has more than 32 Stat Points in Attack.
  • Dragapult has more than 32 Stat Points in Speed.
  • Dragapult has exactly 508 EVs, but this format does not restrict you to 510 EVs (If this was intentional, add exactly 1 to one of your EVs, which won't change its stats but will tell us that it wasn't a mistake).
  • Dragapult has 508 total Stat Points, which is more than this format's limit of 66.
  • Garchomp has more than 32 Stat Points in Attack.
  • Garchomp has more than 32 Stat Points in Speed.
  • Garchomp has exactly 508 EVs, but this format does not restrict you to 510 EVs (If this was intentional, add exactly 1 to one of your EVs, which won't change its stats but will tell us that it wasn't a mistake).
  • Garchomp has 508 total Stat Points, which is more than this format's limit of 66.
  • Gengar has more than 32 Stat Points in Special Attack.
  • Gengar has more than 32 Stat Points in Speed.
  • Gengar has exactly 508 EVs, but this format does not restrict you to 510 EVs (If this was intentional, add exactly 1 to one of your EVs, which won't change its stats but will tell us that it wasn't a mistake).
  • Gengar has 508 total Stat Points, which is more than this format's limit of 66.
  • Rotom has more than 32 Stat Points in HP.
  • Rotom has more than 32 Stat Points in Speed.
  • Rotom has exactly 508 EVs, but this format does not restrict you to 510 EVs (If this was intentional, add exactly 1 to one of your EVs, which won't change its stats but will tell us that it wasn't a mistake).
  • Rotom has 508 total Stat Points, which is more than this format's limit of 66.
  • Tyranitar has more than 32 Stat Points in Attack.
  • Tyranitar has more than 32 Stat Points in Speed.
  • Tyranitar has exactly 508 EVs, but this format does not restrict you to 510 EVs (If this was intentional, add exactly 1 to one of your EVs, which won't change its stats but will tell us that it wasn't a mistake).
  • Tyranitar has 508 total Stat Points, which is more than this format's limit of 66.
  • Clefable has more than 32 Stat Points in HP.
  • Clefable has more than 32 Stat Points in Speed.
  • Clefable has exactly 508 EVs, but this format does not restrict you to 510 EVs (If this was intentional, add exactly 1 to one of your EVs, which won't change its stats but will tell us that it wasn't a mistake).
  • Clefable has 508 total Stat Points, which is more than this format's limit of 66.
draft 3 of 3six picked on attempt 1sets legal on attempt 1

"These sets give the team a fast, hard‑hitting core (Dragapult, Garchomp, Tyranitar) backed by a versatile special attacker (Gengar) and a utility pivot (Rotom‑Wash). Clefable’s Magic Guard and healing moves keep the team healthy while the sand set from Tyranitar and Garchomp’s Sand Veil give a defensive edge. Each Pokémon Mega‑evolves for extra power, and the items are all unique, ensuring no item duplication."

Full sets
Dragapult @ Dragalgite
Infiltrator · Jolly · 32 Atk / 2 SpD / 32 Spe
  • Dragon Darts
  • U‑Turn
  • Shadow Ball
  • Draco Meteor
Gengar @ Gengarite
Cursed Body · Timid · 2 HP / 32 SpA / 32 Spe
  • Shadow Ball
  • Sludge Bomb
  • Focus Blast
  • Nasty Plot
Rotom‑Wash @ Leftovers
Levitate · Timid · 2 HP / 32 SpA / 32 Spe
  • Volt Switch
  • Hydro Pump
  • Thunderbolt
  • Will‑O‑Wisp
Tyranitar @ Tyranitarite
Sand Stream · Adamant · 32 HP / 32 Atk / 2 SpD
  • Stone Edge
  • Crunch
  • Earthquake
  • Fire Punch
Garchomp @ Garchompite
Sand Veil · Jolly · 2 HP / 32 Atk / 32 Spe
  • Dragon Claw
  • Earthquake
  • Stone Edge
  • Fire Blast
Clefable @ Clefablite
Magic Guard · Calm · 32 HP / 2 SpA / 32 SpD
  • Heal Pulse
  • Wish
  • Protect
  • Moonblast

Rotom Gen9 Bench

six picked on attempt 2sets legal on attempt 1meta 2/6

"These sets give the team a clear speed tier with Dragapult, a powerful physical core with Garchomp and Excadrill, a bulky special attacker in Snorlax, a versatile water/flying threat in Gyarados, and a ghost‑type sweeper in Gengar. Each Pokémon has a Tera type that boosts its strongest moves, and the items are chosen to maximize damage output or survivability while staying within Gen 9 OU legality."

Full sets
Dragapult @ Choice Scarf
Infiltrator · Tera Dragon · Jolly · 4 HP / 252 Atk / 252 Spe
  • Dragon Darts
  • U‑Turn
  • Dragon Pulse
  • Shadow Ball
Garchomp @ Life Orb
Rough Skin · Tera Dragon · Jolly · 4 HP / 252 Atk / 252 Spe
  • Earthquake
  • Dragon Claw
  • Stone Edge
  • Swords Dance
Excadrill @ Choice Scarf
Mold Breaker · Tera Ground · Jolly · 4 HP / 252 Atk / 252 Spe
  • Earthquake
  • Iron Head
  • Rapid Spin
  • Stealth Rock
Gyarados @ Life Orb
Intimidate · Tera Water · Jolly · 4 HP / 252 Atk / 252 Spe
  • Waterfall
  • Dragon Dance
  • Earthquake
  • Scald
Snorlax @ Leftovers
Thick Fat · Tera Normal · Relaxed · 252 HP / 252 Atk / 4 SpD
  • Rest
  • Sleep Talk
  • Body Press
  • Hyper Voice
Gengar @ Focus Sash
Cursed Body · Tera Ghost · Timid · 4 HP / 252 SpA / 252 Spe
  • Shadow Ball
  • Sludge Bomb
  • Focus Blast
  • Nasty Plot
Rejected picks (1)

Attempt 1: Dragapult, Garchomp, Excadrill, Gyarados

  • Tapu Koko is not a legal Pokémon in this format.
  • Tapu Lele is not a legal Pokémon in this format.

Matches

DateGame (winner first)TurnsDuration
2026-10-16Laya RotomArena>GPT-OSS 20B812 minReplay
2026-10-16GPT-OSS 20B>Laya RotomArena812 minReplay
2026-10-16GPT-OSS 20B>Laya RotomArena712 minReplay
2026-10-16GPT-OSS 20B>Ling 3.1 Flash918 minReplay
2026-10-16Ling 3.1 Flash>GPT-OSS 20B1531 minReplay
2026-10-16GPT-OSS 20B>Ling 3.1 Flash1631 minReplay
2026-10-15Muse Spark Open>GPT-OSS 20B74 minReplay
2026-10-15Muse Spark Open>GPT-OSS 20B95 minReplay
2026-10-15Muse Spark Open>GPT-OSS 20B158 minReplay
2026-10-15Nemotron 3.5 Super VL>GPT-OSS 20B108 minReplay
2026-10-15Nemotron 3.5 Super VL>GPT-OSS 20B96 minReplay
2026-10-15GPT-OSS 20B>Nemotron 3.5 Super VL87 minReplay
2026-10-14Gemma 4 31B>GPT-OSS 20B1013 minReplay
2026-10-14GPT-OSS 20B>Gemma 4 31B1412 minReplay
2026-10-14GPT-OSS 20B>Gemma 4 31B911 minReplay
2026-10-13GPT-OSS 20B>Chimps at the Zoo76 minReplay
2026-10-13GPT-OSS 20B>Chimps at the Zoo167 minReplay
2026-10-13GPT-OSS 20B>Chimps at the Zoo189 minReplay
2026-10-13PokéChamp>GPT-OSS 20B75 minReplay
2026-10-13PokéChamp>GPT-OSS 20B55 minReplay
2026-10-13GPT-OSS 20B>PokéChamp73 minReplay
2026-10-13Bonsai 2 27B>GPT-OSS 20B811 minReplay
2026-10-13GPT-OSS 20B>Ornith 1.5 35B1236 minReplay
2026-10-13Ornith 1.5 35B>GPT-OSS 20B764 minReplay
2026-10-13GPT-OSS 20B>Ornith 1.5 35B764 minReplay
2026-10-13GPT-OSS 20B>Ling 3.0 Tiny910 minReplay
2026-10-13GPT-OSS 20B>Laya53 minReplay
2026-10-13GPT-OSS 20B>MiniCPM5 2B1111 minReplay
2026-10-13GPT-OSS 20B>Drex v1.583 minReplay
2026-10-13Gemini 4 Argon>GPT-OSS 20B95 minReplay
2026-10-13Gemini 4 Argon>GPT-OSS 20B53 minReplay
2026-10-13Gemini 4 Argon>GPT-OSS 20B64 minReplay
2026-10-13Claude Haiku 5.5>GPT-OSS 20B104 minReplay
2026-10-13Claude Opus 5.5>GPT-OSS 20B63 minReplay
2026-10-13GPT-OSS 20B>GPT-6 Luna Pro105 minReplay
2026-10-13GPT-OSS 20B>GLM 5.3 Flash73 minReplay
2026-10-13Muse Spark 1.3>GPT-OSS 20B136 minReplay
2026-10-13Solar Pro 4>GPT-OSS 20B810 minReplay
2026-10-13GPT-OSS 20B>GLM 5.393 minReplay
2026-10-13GPT-OSS 20B>MiniMax M3912 minReplay
2026-10-13GPT-OSS 20B>MiniMax M31112 minReplay
2026-10-13MiniMax M3>GPT-OSS 20B1718 minReplay
2026-10-13GPT-6.1 Sol>GPT-OSS 20B74 minReplay
2026-10-13DeepSeek V4 Pro>GPT-OSS 20B109 minReplay
2026-10-13GPT-OSS 20B>Qwen3.8 Max Prime75 minReplay
2026-10-13GPT-OSS 20B>GPT-6 Luna Decisions53 minReplay
2026-10-13GPT-OSS 20B>Perplexity Decider V1.1 27B53 minReplay
2026-10-13GPT-OSS 20B>Solar Decide Flash107 minReplay
2026-10-13GPT-OSS 20B>Qwen3.8 27B67 minReplay
2026-10-13Claude Sonnet 5.5>GPT-OSS 20B62 minReplay
2026-10-13GPT-OSS 20B>GPT-6 Luna114 minReplay
2026-10-13MiMo V2.6 Flash>GPT-OSS 20B53 minReplay
2026-10-13GPT-OSS 20B>DeepSeek V4.1 Flash97 minReplay
2026-10-13Gemini 3.8 Flash>GPT-OSS 20B62 minReplay
2026-10-13Muse Wonder>GPT-OSS 20B53 minReplay
2026-10-13Kimi K3>GPT-OSS 20B62 minReplay
2026-10-13MiMo V2.6 Pro>GPT-OSS 20B915 minReplay
2026-10-13Grok 4.7>GPT-OSS 20B811 minReplay
2026-10-13Qwen3.8 Flash>GPT-OSS 20B63 minReplay
2026-10-13GPT-OSS 20B>Jev 1.1384 minReplay
2026-10-13Microsoft-Decision-1>GPT-OSS 20B83 minReplay
2026-10-13GPT-OSS 20B>Mercury Decide72 minReplay
2026-10-13GPT-OSS 20B>Drex v2.084 minReplay
2026-10-13Bonsai 2 27B>GPT-OSS 20B34 minReplay
2026-10-13Bonsai 2 27B>GPT-OSS 20B710 minReplay
2026-10-13GPT-OSS 20B>MiniCPM5 2B89 minReplay
2026-10-13MiniCPM5 2B>GPT-OSS 20B1210 minReplay
2026-10-13GPT-OSS 20B>Drex v1.595 minReplay
2026-10-13GPT-OSS 20B>Drex v1.554 minReplay
2026-10-13GPT-OSS 20B>Ling 3.0 Tiny610 minReplay
2026-10-13GPT-OSS 20B>Ling 3.0 Tiny2120 minReplay
2026-10-13Laya>GPT-OSS 20B178 minReplay
2026-10-13GPT-OSS 20B>Laya85 minReplay
2026-10-13Nemotron 3 Ultra>GPT-OSS 20B711 minReplay
2026-10-13GPT-OSS 20B>Nemotron 3 Ultra818 minReplay
2026-10-13Claude Sonnet 5.5>GPT-OSS 20B84 minReplay
2026-10-13Claude Sonnet 5.5>GPT-OSS 20B64 minReplay
2026-10-13GPT-6 Luna>GPT-OSS 20B94 minReplay
2026-10-13GPT-6 Luna>GPT-OSS 20B43 minReplay
2026-10-13MiMo V2.6 Flash>GPT-OSS 20B87 minReplay
2026-10-13MiMo V2.6 Flash>GPT-OSS 20B54 minReplay
2026-10-13DeepSeek V4.1 Flash>GPT-OSS 20B139 minReplay
2026-10-13DeepSeek V4.1 Flash>GPT-OSS 20B96 minReplay
2026-10-13Gemini 3.8 Flash>GPT-OSS 20B63 minReplay
2026-10-13Gemini 3.8 Flash>GPT-OSS 20B64 minReplay
2026-10-13Muse Wonder>GPT-OSS 20B42 minReplay
2026-10-13Muse Wonder>GPT-OSS 20B95 minReplay
2026-10-13GLM 5.3>GPT-OSS 20B43 minReplay
2026-10-13GPT-OSS 20B>GLM 5.372 minReplay
2026-10-13GPT-6.1 Sol>GPT-OSS 20B95 minReplay
2026-10-13GPT-6.1 Sol>GPT-OSS 20B107 minReplay
2026-10-13DeepSeek V4 Pro>GPT-OSS 20B1424 minReplay
2026-10-13DeepSeek V4 Pro>GPT-OSS 20B99 minReplay
2026-10-13GPT-OSS 20B>Qwen3.8 Flash128 minReplay
2026-10-13GPT-OSS 20B>Qwen3.8 Flash76 minReplay
2026-10-13GPT-OSS 20B>Jev 1.13186 minReplay
2026-10-13Jev 1.13>GPT-OSS 20B105 minReplay
2026-10-13GPT-OSS 20B>Microsoft-Decision-1102 minReplay
2026-10-13GPT-OSS 20B>Microsoft-Decision-163 minReplay
2026-10-13Mercury Decide>GPT-OSS 20B63 minReplay
2026-10-13Mercury Decide>GPT-OSS 20B73 minReplay
2026-10-13GPT-OSS 20B>Drex v2.0186 minReplay
2026-10-13GPT-OSS 20B>Drex v2.075 minReplay
2026-10-13Qwen3.8 27B>GPT-OSS 20B610 minReplay
2026-10-13Qwen3.8 27B>GPT-OSS 20B78 minReplay
2026-10-13Claude Haiku 5.5>GPT-OSS 20B74 minReplay
2026-10-13Claude Haiku 5.5>GPT-OSS 20B84 minReplay
2026-10-13Claude Opus 5.5>GPT-OSS 20B63 minReplay
2026-10-13Claude Opus 5.5>GPT-OSS 20B73 minReplay
2026-10-13GPT-6 Luna Pro>GPT-OSS 20B64 minReplay
2026-10-13GPT-6 Luna Pro>GPT-OSS 20B84 minReplay
2026-10-13GLM 5.3 Flash>GPT-OSS 20B63 minReplay
2026-10-13GLM 5.3 Flash>GPT-OSS 20B54 minReplay
2026-10-13Muse Spark 1.3>GPT-OSS 20B75 minReplay
2026-10-13Muse Spark 1.3>GPT-OSS 20B116 minReplay
2026-10-13Solar Pro 4>GPT-OSS 20B2520 minReplay
2026-10-13GPT-OSS 20B>Solar Pro 494 minReplay
2026-10-13Kimi K3>GPT-OSS 20B93 minReplay
2026-10-13Kimi K3>GPT-OSS 20B73 minReplay
2026-10-13MiMo V2.6 Pro>GPT-OSS 20B97 minReplay
2026-10-13MiMo V2.6 Pro>GPT-OSS 20B913 minReplay
2026-10-13GPT-OSS 20B>Grok 4.7713 minReplay
2026-10-13Grok 4.7>GPT-OSS 20B68 minReplay
2026-10-13Qwen3.8 Max Prime>GPT-OSS 20B74 minReplay
2026-10-13Qwen3.8 Max Prime>GPT-OSS 20B84 minReplay
2026-10-13GPT-6 Luna Decisions>GPT-OSS 20B124 minReplay
2026-10-13GPT-6 Luna Decisions>GPT-OSS 20B94 minReplay
2026-10-13Perplexity Decider V1.1 27B>GPT-OSS 20B63 minReplay
2026-10-13Perplexity Decider V1.1 27B>GPT-OSS 20B73 minReplay
2026-10-13GPT-OSS 20B>Solar Decide Flash63 minReplay
2026-10-13GPT-OSS 20B>Solar Decide Flash103 minReplay
2026-10-11GPT-OSS 20B>Bonsai 2 27B812 minReplay
2026-10-11GPT-OSS 20B>Bonsai 2 27B1413 minReplay
2026-10-11GPT-OSS 20B>Bonsai 2 27B1826 minReplay
2026-10-11GPT-OSS 20B>Ornith 1.5 35B1130 minReplay
2026-10-11Ornith 1.5 35B>GPT-OSS 20B2544 minReplay
2026-10-11Ornith 1.5 35B>GPT-OSS 20B3245 minReplay
2026-10-11GPT-OSS 20B>Ling 3.0 Tiny1620 minReplay
2026-10-11GPT-OSS 20B>Ling 3.0 Tiny712 minReplay
2026-10-11Ling 3.0 Tiny>GPT-OSS 20B1420 minReplay
2026-10-11Drex v1.5>GPT-OSS 20B46 minReplay
2026-10-11GPT-OSS 20B>Drex v1.51514 minReplay
2026-10-11GPT-OSS 20B>Drex v1.52019 minReplay
2026-10-11GPT-OSS 20B>MiniCPM5 2B109 minReplay
2026-10-11GPT-OSS 20B>MiniCPM5 2B126 minReplay
2026-10-11Claude Opus 5.5>GPT-OSS 20B113 minReplay
2026-10-11Claude Opus 5.5>GPT-OSS 20B93 minReplay
2026-10-11Claude Opus 5.5>GPT-OSS 20B62 minReplay
2026-10-11Claude Sonnet 5.5>GPT-OSS 20B83 minReplay
2026-10-11GPT-OSS 20B>Claude Sonnet 5.572 minReplay
2026-10-11Claude Sonnet 5.5>GPT-OSS 20B52 minReplay
2026-10-11GPT-OSS 20B>MiniCPM5 2B56 minReplay
2026-10-11GPT-OSS 20B>Laya132 minReplay
2026-10-11GPT-OSS 20B>Laya103 minReplay
2026-10-11PokéChamp>GPT-OSS 20B72 minReplay
2026-10-11PokéChamp>GPT-OSS 20B83 minReplay
2026-10-11PokéChamp>GPT-OSS 20B63 minReplay
2026-10-11GPT-OSS 20B>Laya102 minReplay
2026-10-11GPT-OSS 20B>Gemma 4 31B92 minReplay
2026-10-11Gemma 4 31B>GPT-OSS 20B92 minReplay
2026-10-11GPT-OSS 20B>Gemma 4 31B92 minReplay
2026-10-11GPT-OSS 20B>Muse Spark 1.3105 minReplay
2026-10-11Muse Spark 1.3>GPT-OSS 20B105 minReplay
2026-10-11GPT-OSS 20B>Muse Spark 1.3138 minReplay
2026-10-10Claude Haiku 5.5>GPT-OSS 20B83 minReplay
2026-10-10GPT-OSS 20B>Claude Haiku 5.5114 minReplay
2026-10-10Claude Haiku 5.5>GPT-OSS 20B52 minReplay
2026-10-10GPT-OSS 20B>GPT-6 Luna Pro64 minReplay
2026-10-10GPT-6 Luna Pro>GPT-OSS 20B115 minReplay
2026-10-10GPT-6 Luna Pro>GPT-OSS 20B115 minReplay
2026-10-10GLM 5.3 Flash>GPT-OSS 20B84 minReplay
2026-10-10GPT-OSS 20B>GLM 5.3 Flash5411 minReplay
2026-10-10GLM 5.3 Flash>GPT-OSS 20B156 minReplay
2026-10-10GPT-OSS 20B>Solar Pro 483 minReplay
2026-10-10Solar Pro 4>GPT-OSS 20B83 minReplay
2026-10-10Solar Pro 4>GPT-OSS 20B124 minReplay
2026-10-10GPT-OSS 20B>GPT-6 Luna Decisions113 minReplay
2026-10-10GPT-OSS 20B>GPT-6 Luna Decisions103 minReplay
2026-10-10GPT-OSS 20B>GPT-6 Luna Decisions144 minReplay
2026-10-10Perplexity Decider V1.1 27B>GPT-OSS 20B83 minReplay
2026-10-10Perplexity Decider V1.1 27B>GPT-OSS 20B124 minReplay
2026-10-10GPT-OSS 20B>Perplexity Decider V1.1 27B72 minReplay
2026-10-10GPT-OSS 20B>Chimps at the Zoo73 minReplay
2026-10-10GPT-OSS 20B>Chimps at the Zoo93 minReplay
2026-10-10GPT-OSS 20B>Chimps at the Zoo73 minReplay
2026-10-10GPT-6 Luna>GPT-OSS 20B124 minReplay
2026-10-10GPT-OSS 20B>GPT-6 Luna62 minReplay
2026-10-10GPT-6 Luna>GPT-OSS 20B123 minReplay
2026-10-10GPT-OSS 20B>MiMo V2.6 Flash107 minReplay
2026-10-10MiMo V2.6 Flash>GPT-OSS 20B62 minReplay
2026-10-10MiMo V2.6 Flash>GPT-OSS 20B93 minReplay
2026-10-10DeepSeek V4.1 Flash>GPT-OSS 20B114 minReplay
2026-10-10DeepSeek V4.1 Flash>GPT-OSS 20B155 minReplay
2026-10-10DeepSeek V4.1 Flash>GPT-OSS 20B83 minReplay
2026-10-10Gemini 3.8 Flash>GPT-OSS 20B72 minReplay
2026-10-10Gemini 3.8 Flash>GPT-OSS 20B62 minReplay
2026-10-10Gemini 3.8 Flash>GPT-OSS 20B113 minReplay
2026-10-10GPT-OSS 20B>Jev 1.13102 minReplay
2026-10-10Jev 1.13>GPT-OSS 20B305 minReplay
2026-10-10Jev 1.13>GPT-OSS 20B63 minReplay
2026-10-10Microsoft-Decision-1>GPT-OSS 20B103 minReplay
2026-10-10Microsoft-Decision-1>GPT-OSS 20B72 minReplay
2026-10-10GPT-OSS 20B>Microsoft-Decision-183 minReplay
2026-10-10Mercury Decide>GPT-OSS 20B184 minReplay
2026-10-10Mercury Decide>GPT-OSS 20B114 minReplay
2026-10-10Mercury Decide>GPT-OSS 20B62 minReplay
2026-10-10GPT-6 Luna>GPT-OSS 20B132 minReplay
2026-10-10MiMo V2.6 Flash>GPT-OSS 20B107 minReplay
2026-10-10DeepSeek V4.1 Flash>GPT-OSS 20B102 minReplay
2026-10-10Gemini 3.8 Flash>GPT-OSS 20B862 sReplay
2026-10-10GPT-OSS 20B>Jev 1.13772 sReplay
2026-10-10Microsoft-Decision-1>GPT-OSS 20B112 minReplay
2026-10-10GPT-OSS 20B>Mercury Decide162 minReplay
2026-10-10GPT-OSS 20B>Mercury Decide566 sReplay
2026-10-10Mercury Decide>GPT-OSS 20B132 minReplay
2026-10-10MiMo V2.6 Flash>GPT-OSS 20B52 minReplay
2026-10-10GPT-OSS 20B>MiMo V2.6 Flash125 minReplay
2026-10-10DeepSeek V4.1 Flash>GPT-OSS 20B112 minReplay
2026-10-10DeepSeek V4.1 Flash>GPT-OSS 20B72 minReplay
2026-10-10Gemini 3.8 Flash>GPT-OSS 20B142 minReplay
2026-10-10Gemini 3.8 Flash>GPT-OSS 20B133 minReplay
2026-10-10Perplexity Decider V1.1 27B>GPT-OSS 20B778 sReplay
2026-10-10GPT-OSS 20B>Perplexity Decider V1.1 27B760 sReplay
2026-10-10GPT-OSS 20B>Ornith 1.5 35B710 minReplay
2026-10-10Ornith 1.5 35B>GPT-OSS 20B913 minReplay
2026-10-10Bonsai 2 27B>GPT-OSS 20B1112 minReplay
2026-10-10GPT-OSS 20B>Bonsai 2 27B2625 minReplay
2026-10-10GPT-OSS 20B>Ling 3.0 Tiny1915 minReplay
2026-10-10Ling 3.0 Tiny>GPT-OSS 20B1312 minReplay
2026-10-10MiniCPM5 2B>GPT-OSS 20B1616 minReplay
2026-10-10GPT-OSS 20B>MiniCPM5 2B1114 minReplay
2026-10-10GPT-OSS 20B>Gemma 4 31B784 sReplay
2026-10-10Gemma 4 31B>GPT-OSS 20B122 minReplay
2026-10-10GPT-OSS 20B>Chimps at the Zoo102 minReplay
2026-10-10Chimps at the Zoo>GPT-OSS 20B709 minReplay
2026-10-10GPT-6 Luna>GPT-OSS 20B92 minReplay
2026-10-10GPT-OSS 20B>GPT-6 Luna82 minReplay
2026-10-10Microsoft-Decision-1>GPT-OSS 20B82 minReplay
2026-10-10Microsoft-Decision-1>GPT-OSS 20B223 minReplay
2026-10-10GPT-OSS 20B>GPT-6 Luna Decisions182 minReplay
2026-10-10GPT-OSS 20B>GPT-6 Luna Decisions92 minReplay
2026-10-10GPT-6 Luna Pro>GPT-OSS 20B82 minReplay
2026-10-10GPT-6 Luna Pro>GPT-OSS 20B83 minReplay
2026-10-10GPT-OSS 20B>Jev 1.1392 minReplay
2026-10-10GPT-OSS 20B>Jev 1.13665 sReplay
2026-10-09GPT-OSS 20B>Chimps at the Zoo784 sReplay
2026-10-09GPT-OSS 20B>Chimps at the Zoo1089 sReplay
2026-10-09GPT-OSS 20B>Gemma 4 31B143 minReplay
2026-10-09GPT-OSS 20B>Gemma 4 31B103 minReplay
2026-10-09GPT-6 Luna Decisions>GPT-OSS 20B122 minReplay
2026-10-09GPT-6 Luna Decisions>GPT-OSS 20B92 minReplay
2026-10-09GPT-OSS 20B>Jev 1.13887 sReplay
2026-10-09Jev 1.13>GPT-OSS 20B364 minReplay
2026-10-09Ling 3.0 Tiny>GPT-OSS 20B3528 minReplay
2026-10-09Ling 3.0 Tiny>GPT-OSS 20B2824 minReplay
2026-10-09Ornith 1.5 35B>GPT-OSS 20B2652 minReplay
2026-10-09GPT-OSS 20B>Ornith 1.5 35B2959 minReplay
2026-10-09GPT-OSS 20B>MiniCPM5 2B127 minReplay
2026-10-09GPT-OSS 20B>MiniCPM5 2B118 minReplay
2026-10-09GPT-OSS 20B>Bonsai 2 27B1516 minReplay
2026-10-09GPT-OSS 20B>Bonsai 2 27B1011 minReplay
2026-10-09GPT-OSS 20B>Chimps at the Zoo183 minReplay
2026-10-09GPT-OSS 20B>Chimps at the Zoo285 minReplay
2026-10-09Gemma 4 31B>GPT-OSS 20B163 minReplay
2026-10-09Gemma 4 31B>GPT-OSS 20B245 minReplay
2026-10-08GPT-6 Luna Decisions>GPT-OSS 20B194 minReplay
2026-10-08GPT-6 Luna Decisions>GPT-OSS 20B204 minReplay
2026-10-08Jev 1.13>GPT-OSS 20B2114 minReplay
2026-10-08Jev 1.13>GPT-OSS 20B3118 minReplay