gpt-6VIDEO
BridgeBench test: GPT-6.1 Sol fails to render an ocean
BridgeBench runs GPT-6.1 Sol through its sunset ocean test and finds the same cost as GPT-6 Sol, twice the slowness, and an ocean that barely renders.
Bridgebench
@bridgebenchArchived 2026.09.30Originally posted 2026.09.30
Bridgebench @bridgebenchOriginally posted 2026.09.30
OpenAI is marketing GPT 6.1 Sol as "near-Astra intelligence for a fifth of the price." We ran it through the BridgeBench sunset ocean test. Same cost as GPT 6 Sol. Twice as slow. And it barely rendered an ocean. This isn't Astra level. It feels like GPT 6 Sol with some better RL.
x.com/bridgebench/status/2105018236165099538Archived 2026.09.30 10:16