Model profile

GPT-6 Astra playing video games

8 GAMES

Games played: Chess · Go (board game) · Heads-up Poker · Minecraft · NetHack · NetHack & roguelikes (BALROG) · Pokémon (game not stated) · Pokémon VGC (competitive doubles). Each game below is scored on its own terms.

Achievements

Achievements
Model	Game	Achievement	Run config	Status	Date
GPT-6 Astra reasoning effort: high	Pokémon (game not stated)	Became Champion in 18h 12m, the fastest claimed AI Champion time we know of Last verified: not independently verified · Last checked: Oct 18, 2026 How was this run measured? Model GPT-6 Astra (source label: “GPT-6 Astra (high)”) Game Pokémon (game not stated) Project GPT Plays Pokémon Harness GPT Plays Pokémon vision-only harness Inputs vision-only Tools custom tools (not itemised) Assistance not disclosed Budget reasoning effort: high Verification Creator-reported Creator's own thread. The game version is not stated there. Evidence Creator thread (unrolled) (creator) GPT Plays Pokémon on Twitch (creator) Unknown which Pokémon game; date completed Dates evidence dated Sep 5, 2026 Recorded published Sep 5, 2026 · discovered by AI vs Games Oct 17, 2026 Changes Historical record. Nothing to update unless the source changes. Permanent link · Submit a correction · Share card (PNG) · JSON	GPT Plays Pokémon vision-only harness vision-only	Creator-reported source	Sep 5, 2026
GPT-6 Astra reasoning effort: medium	Minecraft	Top Mine Odyssey score: 85.6% of 180 navigation tasks in Minecraft rebuilds of real places Claude Opus 5.5 (max) was second at 73.9%. The best open-weight model managed 23.9%. Last verified: not independently verified · Last checked: Oct 18, 2026 How was this run measured? Model GPT-6 Astra (source label: “GPT-6 Astra (medium)”) Game Minecraft Project Mine Odyssey Harness Mine Odyssey navigation harness Inputs not disclosed Tools not disclosed Assistance none Budget reasoning effort: medium Verification Creator-reported Paper authors' results; no public leaderboard yet. Evidence Mine Odyssey paper (arXiv 2610.11328), Table 1 (creator) Dates achieved Oct 8, 2026 · evidence dated Oct 8, 2026 Recorded published Oct 8, 2026 · discovered by AI vs Games Oct 18, 2026 Changes Historical record. Nothing to update unless the source changes. Permanent link · Submit a correction · JSON	Mine Odyssey navigation harness not disclosed	Creator-reported source	Oct 8, 2026
GPT-6 Astra	Go	#1 on Kaggle Game Arena Go: 1794 Elo Last verified: – · Last checked: Oct 18, 2026 How was this run measured? Model GPT-6 Astra Game Go Project Kaggle Game Arena Harness Kaggle Game Arena harness Inputs board as text Tools none Assistance none Budget per-model settings not shown Verification Verified Read from the official leaderboard, which links replays of every game. Evidence Kaggle Game Arena: Go leaderboard (official) (operator) Dates evidence dated Oct 15, 2026 Recorded published Oct 15, 2026 · discovered by AI vs Games Oct 18, 2026 Changes No change detected since we started tracking. Permanent link · Submit a correction · Share card (PNG) · JSON	Kaggle Game Arena harness board as text	Verified source	Oct 15, 2026

Memorable failures

Failures
Model	Game	Achievement	Run config	Status	Date
GPT-6 Astra reasoning effort: max	NetHack	Memorable fail 13.2% progress on NetHack for the top overall BALROG entry. Nobody has beaten the game. Last verified: Oct 18, 2026 · Last checked: Oct 18, 2026 How was this run measured? Model GPT-6 Astra (source label: “GPT-6-Astra-Max”) Game NetHack Project BALROG Harness BALROG agentic harness Inputs text / vision per track Tools none Assistance none Budget reasoning effort: max Verification Verified Read from the official leaderboard. Evidence BALROG leaderboard (official) (operator) Dates evidence dated Sep 18, 2026 Recorded published Sep 18, 2026 · discovered by AI vs Games Oct 17, 2026 Changes No change detected since we started tracking. Permanent link · Submit a correction · JSON	BALROG agentic harness text / vision per track	Verified source	Sep 18, 2026

Leaderboard placements one row per game, never combined

Game / board	Rank	Result	Status	As of
BALROG: BabyAI, Crafter, TextWorld, BabaIsAI, MiniHack, NetHack as GPT-6 Astra (max)	1	68.3 Avg progress %	Verified	2026-09-18
Kaggle Game Arena: Chess (text) as GPT-6 Astra	2	1,421 Elo	Verified	2026-10-15
Kaggle Game Arena: Go as GPT-6 Astra	1	1,794 Elo	Verified	2026-10-15
Kaggle Game Arena: Heads-up Poker as GPT-6 Astra	3	25 Mean BB/100	Verified	2026-10-08
Pokémon VGC Bench as GPT-6 Astra	1	1,901 Rating	Verified	2026-10-04

Model versions seen

GPT-6 Astra · 6 GPT-6 Astra (max) · 2 GPT-6 Astra (high) · 1 GPT-6 Astra (medium) · 1

Milestones timeline

Sep 5, 2026
Pokémon (game not stated): Became Champion in 18h 12m, the fastest claimed AI Champion time we know of Creator-reported source
Sep 18, 2026
✗ NetHack: 13.2% progress on NetHack for the top overall BALROG entry. Nobody has beaten the game. Verified source
Oct 8, 2026
Minecraft: Top Mine Odyssey score: 85.6% of 180 navigation tasks in Minecraft rebuilds of real places Creator-reported source
Oct 15, 2026
Go: #1 on Kaggle Game Arena Go: 1794 Elo Verified source

This Week in AI vs Games

One short email every Sunday: new records, memorable failures and runs to watch. Free. Every claim is linked to its source.

Latest issue: Issue #1: The Ender Dragon is down