Claude Haiku 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 75.3)
- •DeepSeek V3.2 Speciale is cheaper ($0.29 vs $1.00 per 1M input)
- •Claude Haiku 4.5 has a larger context window (200K)
| Claude Haiku 4.5 | DeepSeek V3.2 Speciale | |
|---|---|---|
| Intelligence index | 75.3 | 89.9 |
| Developer | Anthropic | DeepSeek |
| Type | LLM | LLM |
| Access | API only | Open weights |
| Context window | 200,000 tokens | 163,840 tokens |
| Input price | $1.00 / 1M | $0.29 / 1M |
| Output price | $5.00 / 1M | $0.43 / 1M |
| Speed | 100 tok/s | — |
| Released | October 15, 2025 | December 1, 2025 |
| Parameters | — | — |
| Input modalities | Text, Image | Text |
| Output modalities | Text | Text |
Shared benchmarks
Claude Haiku 4.5
DeepSeek V3.2 Speciale
AIME 2025
96.3
96.7
GPQA Diamond
73
87.1
Humanity’s Last Exam
9.7
26.1
LiveCodeBench
61.5
89.6
MMLU-Pro
80
86.3