gemini-4THREAD
Gemini Argon 4 scores 12% on MazeBench for second place
Gemini Argon 4 (apparently Gemini 4 Pro) reportedly scores 12% on MazeBench, placing second just below GPT-6 Astra.
💺
@patience_caveArchived 21 days agoOriginally posted 2026.10.15
THREAD
MazeBench vs Gemini Argon 4 There’s a new kid on the block Gemini Argon 4 scores 12% It’s a 2nd place score, just below GPT-6 Astra

Gemini Argon 4 explores similar rooms to GPT-6 Astra. It is slightly more thorough, discovering gems that Astra overlooked. GPT-6 Astra is more capable, but also more sporadic in exploration.

💺 @patience_caveOriginally posted 2026.10.15
x.com/patience_cave/status/2110749850010812886Archived 2026.10.16 10:10