0

DeepSeek V3.2 vs GPT-5.4 Mini

Head to head on 22 shared benchmarks, with price, context window, and release dates.

Benchmark scores

41

evals · sort by any column, ⤢ to expand fullscreen

Model
DeepSeek V3.2

DeepSeek

47.9%1511elo66.822.1%75.1%11.2%49.0%68.0%65.9%64.6%50.0%48.2%70.4%85.0%77.2%8.0%38.7%68.2%32.6%655.1%78.9%74.7%74.2%59.0%-0.0%16.0%---31.3%59.3%57.5%83.7%--59.0%70.0%--51.7%
GPT-5.4 Mini

OpenAI

60.9%1662elo7328.3%60.6%5.9%38.8%80.3%70.9%71.6%70.9%60.3%71.5%78.6%72.5%21.0%39.6%71.2%18.2%61.748.0%23.4%---37.0%--21.2%3.9%45.4%----63.5%50.8%--39.6%54.2%48.4%
2 / 2 models
AttributeDeepSeek V3.2

DeepSeek

GPT-5.4 Mini

OpenAI

DescriptionDeepSeek V3.2 is an AI model from DeepSeek, released with open weights.GPT-5.4 mini is an AI model from OpenAI.
FamilyDeepseekGPT
Params--
Open weightsopenclosed
Licensemitproprietary
Released1 Dec 202517 Mar 2026
Context163,840400,000
Input $/1M$0.28$0.75
Output $/1M$0.42$4.50
Cache read $/1M$0.130$0.075
Cache write $/1M--
Speed0 tok/s0 tok/s
Latency0.00 s0.00 s
Modalitiestextfile, image, text
Providers--
PublisherDeepSeekOpenAI
Reported scores
GPQA Diamond60.6% (AA)
IFBench38.8% (AA)
SciCode39.6% (AA)

FAQ

Is DeepSeek V3.2 or GPT-5.4 Mini better?
Across the 22 benchmarks both models report on Sophon, GPT-5.4 Mini leads on 14 of them. Which one is "better" depends on the benchmark - the table above shows every shared score.
How do DeepSeek V3.2 and GPT-5.4 Mini compare on GPQA Diamond?
DeepSeek V3.2 scores 75.1% and GPT-5.4 Mini scores 60.6% on GPQA Diamond.
Which is cheaper, DeepSeek V3.2 or GPT-5.4 Mini?
DeepSeek V3.2 is cheaper at $0.42 per million output tokens against $4.50 for GPT-5.4 Mini - about 10.7x.
When were DeepSeek V3.2 and GPT-5.4 Mini released?
DeepSeek V3.2 was released 1 Dec 2025 by DeepSeek. GPT-5.4 Mini was released 17 Mar 2026 by OpenAI.

Comparing something else? Build your own side-by-side.