Compare AI Models

Comparing the specifications and intelligence benchmarks for DeepSeek V3.1 Terminus (Reasoning), DeepSeek V3.2 (Reasoning).

Model	Intelligence	MMLU Pro	Context	GPQA	TTFT (s)	Speed (T/s)	Input Price (1M)	Output Price (1M)
DeepSeek V3.1 Terminus (Reasoning) DeepSeek	33.9	85.1%	-	79.2%	-	-	$0.40	$2.00
DeepSeek V3.2 (Reasoning) DeepSeek	41.7	86.2%	-	84.0%	1.42	29	$0.28	$0.42