Side-by-side benchmark comparison across agentic, coding, multimodal, knowledge, reasoning, and math workflows.
GPT-5.2 Pro and GPT-5.4 finish on the same overall score, so this is less about a single winner and more about where the edge shows up. The headline says tie; the benchmark table is where the real choice happens.
GPT-5.4's sharpest advantage is in mathematics, where it averages 98.2 against 98.2. The single biggest benchmark swing on the page is SWE-bench Pro, 89 to 85. GPT-5.2 Pro does hit back in agentic, so the answer changes if that is the part of the workload you care about most.
GPT-5.2 Pro is also the more expensive model on tokens at $25.00 input / $150.00 output per 1M tokens, versus $2.50 input / $15.00 output per 1M tokens for GPT-5.4. That is roughly 10.0x on output cost alone. GPT-5.4 gives you the larger context window at 1.05M, compared with 400K for GPT-5.2 Pro.
Treat this as a split decision. GPT-5.2 Pro makes more sense if multimodal & grounded is the priority; GPT-5.4 is the better fit if agentic is the priority or you want the cheaper token bill.
GPT-5.2 Pro
85.9
GPT-5.4
87.8
GPT-5.2 Pro
84.8
GPT-5.4
84.5
GPT-5.2 Pro
96
GPT-5.4
95.5
GPT-5.2 Pro
95.2
GPT-5.4
95.8
GPT-5.2 Pro
81.5
GPT-5.4
82.6
GPT-5.2 Pro
95
GPT-5.4
96
GPT-5.2 Pro
93.4
GPT-5.4
94.7
GPT-5.2 Pro
98.2
GPT-5.4
98.2
GPT-5.2 Pro and GPT-5.4 are tied on overall score, so the right pick depends on which category matters most for your use case.
GPT-5.4 has the edge for knowledge tasks in this comparison, averaging 82.6 versus 81.5. Inside this category, HLE is the benchmark that creates the most daylight between them.
GPT-5.2 Pro has the edge for coding in this comparison, averaging 84.8 versus 84.5. Inside this category, SWE-bench Pro is the benchmark that creates the most daylight between them.
GPT-5.2 Pro and GPT-5.4 are effectively tied for math here, both landing at 98.2 on average.
GPT-5.4 has the edge for reasoning in this comparison, averaging 95.8 versus 95.2. Inside this category, LongBench v2 is the benchmark that creates the most daylight between them.
GPT-5.4 has the edge for agentic tasks in this comparison, averaging 87.8 versus 85.9. Inside this category, OSWorld-Verified is the benchmark that creates the most daylight between them.
GPT-5.2 Pro has the edge for multimodal and grounded tasks in this comparison, averaging 96 versus 95.5. Inside this category, MMMU-Pro is the benchmark that creates the most daylight between them.
GPT-5.4 has the edge for instruction following in this comparison, averaging 96 versus 95. Inside this category, IFEval is the benchmark that creates the most daylight between them.
GPT-5.4 has the edge for multilingual tasks in this comparison, averaging 94.7 versus 93.4. Inside this category, MMLU-ProX is the benchmark that creates the most daylight between them.
Get notified when new models drop, benchmark scores change, or the leaderboard shifts. One email per week.
Free. No spam. Unsubscribe anytime. We only store derived location metadata for consent routing.