Head-to-head comparison across 1benchmark categories. Overall scores shown here use BenchLM's provisional ranking lane.
Claude 3.5 Sonnet
41
o1
58
Pick o1 if you want the stronger benchmark profile. Claude 3.5 Sonnet only becomes the better choice if you want the cheaper token bill or you would rather avoid the extra latency and token burn of a reasoning model.
Knowledge
+16.3 difference
Claude 3.5 Sonnet
o1
$3 / $15
$15 / $60
N/A
98 t/s
N/A
32.29s
200K
200K
Pick o1 if you want the stronger benchmark profile. Claude 3.5 Sonnet only becomes the better choice if you want the cheaper token bill or you would rather avoid the extra latency and token burn of a reasoning model.
o1 is clearly ahead on the provisional aggregate, 58 to 41. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.
o1's sharpest advantage is in knowledge, where it averages 75.7 against 59.4. The single biggest benchmark swing on the page is GPQA, 59.4% to 75.7%.
o1 is also the more expensive model on tokens at $15.00 input / $60.00 output per 1M tokens, versus $3.00 input / $15.00 output per 1M tokens for Claude 3.5 Sonnet. That is roughly 4.0x on output cost alone. o1 is the reasoning model in the pair, while Claude 3.5 Sonnet 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.
o1 is ahead on BenchLM's provisional leaderboard, 58 to 41. The biggest single separator in this matchup is GPQA, where the scores are 59.4% and 75.7%.
o1 has the edge for knowledge tasks in this comparison, averaging 75.7 versus 59.4. Inside this category, GPQA is the benchmark that creates the most daylight between them.
For engineers, researchers, and the plain curious — a weekly brief on new models, ranking shifts, and pricing changes.
Free. No spam. Unsubscribe anytime.