AI Hub
← All models

DeepSeek Coder V2 Lite Instruct vs DeepSeek V3.2 Speciale

DeepSeek vs DeepSeek — benchmarks, pricing, and capabilities side by side.

  • DeepSeek V3.2 Speciale has the higher intelligence index (89.9 vs 30.2)
  • DeepSeek Coder V2 Lite Instruct is cheaper ($0.00 vs $0.29 per 1M input)
DeepSeek Coder V2 Lite InstructDeepSeek V3.2 Speciale
Intelligence index30.289.9
DeveloperDeepSeekDeepSeek
TypeLLMLLM
AccessOpen weights
Context window163,840 tokens
Input price$0.00 / 1M$0.29 / 1M
Output price$0.00 / 1M$0.43 / 1M
Speed
ReleasedJune 17, 2024December 1, 2025
Parameters
Input modalitiesText
Output modalitiesText

Shared benchmarks

DeepSeek Coder V2 Lite Instruct
DeepSeek V3.2 Speciale
GPQA Diamond
31.9
87.1
Humanity’s Last Exam
5.3
26.1
LiveCodeBench
15.8
89.6
MMLU-Pro
42.9
86.3