AI Hub
← All models

Claude Sonnet 4.5 vs DeepSeek V3.2 Speciale

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

  • DeepSeek V3.2 Speciale has the higher intelligence index (89.9 vs 80.4)
  • DeepSeek V3.2 Speciale is cheaper ($0.29 vs $3.00 per 1M input)
  • Claude Sonnet 4.5 has a larger context window (1M)
Claude Sonnet 4.5DeepSeek V3.2 Speciale
Intelligence index80.489.9
DeveloperAnthropicDeepSeek
TypeLLMLLM
AccessAPI onlyOpen weights
Context window1,000,000 tokens163,840 tokens
Input price$3.00 / 1M$0.29 / 1M
Output price$15.00 / 1M$0.43 / 1M
Speed42 tok/s
ReleasedSeptember 29, 2025December 1, 2025
Parameters
Input modalitiesText, ImageText
Output modalitiesTextText

Shared benchmarks

Claude Sonnet 4.5
DeepSeek V3.2 Speciale
AIME 2025
87
96.7
GPQA Diamond
83.4
87.1
Humanity’s Last Exam
17.3
26.1
LiveCodeBench
71.4
89.6
MMLU-Pro
87.5
86.3