Claude Sonnet 4.6 vs GPT-5.5 Pro

Pick GPT-5.5 Pro if you want the stronger benchmark profile. Claude Sonnet 4.6 only becomes the better choice if knowledge is the priority or you want the cheaper token bill.

GPT-5.5 Pro is clearly ahead on the provisional aggregate, 100 to 85. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.

GPT-5.5 Pro's sharpest advantage is in agentic, where it averages 90.1 against 65.3. The single biggest benchmark swing on the page is HLE, 49% to 57.2%. Claude Sonnet 4.6 does hit back in knowledge, so the answer changes if that is the part of the workload you care about most.

GPT-5.5 Pro is also the more expensive model on tokens at $30.00 input / $180.00 output per 1M tokens, versus $3.00 input / $15.00 output per 1M tokens for Claude Sonnet 4.6. That is roughly 12.0x on output cost alone. GPT-5.5 Pro is the reasoning model in the pair, while Claude Sonnet 4.6 is not. That usually helps on harder chain-of-thought-heavy tests, but it can also mean more latency and more token spend in real use. GPT-5.5 Pro gives you the larger context window at 1M, compared with 200K for Claude Sonnet 4.6.