Head-to-head comparison across 3benchmark categories. Overall scores shown here use BenchLM's provisional ranking lane.
GPT-5.2
78
Grok 4.3
74
Pick GPT-5.2 if you want the stronger benchmark profile. Grok 4.3 only becomes the better choice if you want the cheaper token bill or you need the larger 1M context window.
Coding
+17.4 difference
Knowledge
+38.5 difference
Multimodal
+2.2 difference
GPT-5.2
Grok 4.3
$1.75 / $14
$1.25 / $2.5
73 t/s
209 t/s
130.34s
12.36s
400K
1M
Pick GPT-5.2 if you want the stronger benchmark profile. Grok 4.3 only becomes the better choice if you want the cheaper token bill or you need the larger 1M context window.
GPT-5.2 is clearly ahead on the provisional aggregate, 78 to 74. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.
GPT-5.2's sharpest advantage is in knowledge, where it averages 92.4 against 53.9. The single biggest benchmark swing on the page is GPQA, 92.4% to 90.1%.
GPT-5.2 is also the more expensive model on tokens at $1.75 input / $14.00 output per 1M tokens, versus $1.25 input / $2.50 output per 1M tokens for Grok 4.3. That is roughly 5.6x on output cost alone. Grok 4.3 gives you the larger context window at 1M, compared with 400K for GPT-5.2.
GPT-5.2 is ahead on BenchLM's provisional leaderboard, 78 to 74. The biggest single separator in this matchup is GPQA, where the scores are 92.4% and 90.1%.
GPT-5.2 has the edge for knowledge tasks in this comparison, averaging 92.4 versus 53.9. Inside this category, AA-Omniscience Hallucination Rate is the benchmark that creates the most daylight between them.
GPT-5.2 has the edge for coding in this comparison, averaging 64.7 versus 47.3. Inside this category, Terminal-Bench Hard is the benchmark that creates the most daylight between them.
GPT-5.2 has the edge for multimodal and grounded tasks in this comparison, averaging 80.3 versus 78.1. Inside this category, Design Arena Website 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.