AI Hub
← All models

DeepSeek V3.2 Speciale vs GLM-4.6

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

  • DeepSeek V3.2 Speciale has the higher intelligence index (89.9 vs 72.4)
  • DeepSeek V3.2 Speciale is cheaper ($0.29 vs $0.43 per 1M input)
  • GLM-4.6 has a larger context window (203K)
DeepSeek V3.2 SpecialeGLM-4.6
Intelligence index89.972.4
DeveloperDeepSeekZhipu AI
TypeLLMLLM
AccessOpen weightsOpen weights
Context window163,840 tokens202,752 tokens
Input price$0.29 / 1M$0.43 / 1M
Output price$0.43 / 1M$1.74 / 1M
Speed85 tok/s
ReleasedDecember 1, 2025September 30, 2025
Parameters357B (MoE)
Input modalitiesTextText
Output modalitiesTextText

Shared benchmarks

DeepSeek V3.2 Speciale
GLM-4.6
AIME 2025
96.7
93.9
GPQA Diamond
87.1
81
Humanity’s Last Exam
26.1
17.2
LiveCodeBench
89.6
69.5
MMLU-Pro
86.3
82.9