Head-to-head comparison across 3benchmark categories. Overall scores shown here use BenchLM's provisional ranking lane.
GPT-5.2
81
Hy3 Preview
62
Pick GPT-5.2 if you want the stronger benchmark profile. Hy3 Preview only becomes the better choice if you want the cheaper token bill.
Agentic
+0.8 difference
Coding
+4.7 difference
Knowledge
+45.7 difference
GPT-5.2
Hy3 Preview
$1.75 / $14
$0 / $0
73 t/s
N/A
130.34s
N/A
400K
256K
Pick GPT-5.2 if you want the stronger benchmark profile. Hy3 Preview only becomes the better choice if you want the cheaper token bill.
GPT-5.2 is clearly ahead on the provisional aggregate, 81 to 62. 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 46.7. The single biggest benchmark swing on the page is SWE-bench Verified, 80% to 74.4%.
GPT-5.2 is also the more expensive model on tokens at $1.75 input / $14.00 output per 1M tokens, versus $0.00 input / $0.00 output per 1M tokens for Hy3 Preview. That is roughly Infinityx on output cost alone. GPT-5.2 gives you the larger context window at 400K, compared with 256K for Hy3 Preview.
GPT-5.2 is ahead on BenchLM's provisional leaderboard, 81 to 62. The biggest single separator in this matchup is SWE-bench Verified, where the scores are 80% and 74.4%.
GPT-5.2 has the edge for knowledge tasks in this comparison, averaging 92.4 versus 46.7. Inside this category, GPQA 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 60. Inside this category, SWE-bench Verified is the benchmark that creates the most daylight between them.
GPT-5.2 has the edge for agentic tasks in this comparison, averaging 55.2 versus 54.4. Hy3 Preview stays close enough that the answer can still flip depending on your workload.
For engineers, researchers, and the plain curious — a weekly brief on new models, ranking shifts, and pricing changes.
Free. No spam. Unsubscribe anytime.