CHESS BENCHMARK · REASONING MEDIUM

Gemini 4 Argon

Google · released 2026-09-30 · route gemini-4-argon · Model profile

Opening gap: +0 (95%: -1,146 to +1,146). The interval includes zero, so we can't yet say whether it performs differently across the variants.

Standard

Rating
1,41495%: 604 to 2,225
Record
1 won, 0 drawn, 0 lost of 1
Accuracy (open positions)
93.8
Accuracy (all moves)
95.7
Average centipawn loss
18
Blunders / mistakes / inaccuracies per 100 moves
0.0 / 0.0 / 0.0
Matched Stockfish's first choice
67%
In Stockfish's top three
87%
Legal on the first try
100.0%
Illegal / invalid
0 of 15 answers
Forfeits
0
Adjudicated games
0
API-equivalent cost per game
$0.089
API-equivalent cost per move
$0.0059
Seconds per move
9.3
Model minutes per game
2.3
Tokens per move (in / out)
2,192 / 152
Reasoning tokens per move
138
Aborted on our side (not counted)
0

By phase

PhaseMovesAccuracyCentipawn lossBlunders /100Engine match
Opening1093.8200.060%
Middlegame5n/a130.080%
Endgame0n/an/an/an/a

Chess960

Rating
1,41495%: 604 to 2,225
Record
1 won, 0 drawn, 0 lost of 1
Accuracy (open positions)
83.2
Accuracy (all moves)
85.7
Average centipawn loss
93
Blunders / mistakes / inaccuracies per 100 moves
4.9 / 8.2 / 9.8
Matched Stockfish's first choice
39%
In Stockfish's top three
61%
Legal on the first try
98.4%
Illegal / invalid
1 of 62 answers
Forfeits
0
Adjudicated games
1
API-equivalent cost per game
$0.518
API-equivalent cost per move
$0.0085
Seconds per move
11.8
Model minutes per game
12.0
Tokens per move (in / out)
2,460 / 357
Reasoning tokens per move
330
Aborted on our side (not counted)
0

By phase

PhaseMovesAccuracyCentipawn lossBlunders /100Engine match
Opening1089.2400.050%
Middlegame2883.01043.636%
Endgame2379.21028.739%

Games