Apertus 70B Instruct vs DeepSeek-V4-Flash
Swiss AI Initiative vs DeepSeek — benchmarks, pricing, and capabilities side by side.
- •DeepSeek-V4-Flash has the higher intelligence index (89.4 vs 27.2)
- •DeepSeek-V4-Flash is cheaper ($0.10 vs $0.80 per 1M input)
| Apertus 70B Instruct | DeepSeek-V4-Flash | |
|---|---|---|
| Intelligence index | 27.2 | 89.4 |
| Developer | Swiss AI Initiative | DeepSeek |
| Type | LLM | LLM |
| Access | — | Open weights |
| Context window | — | 1,048,576 tokens |
| Input price | $0.80 / 1M | $0.10 / 1M |
| Output price | $2.90 / 1M | $0.20 / 1M |
| Speed | — | 109 tok/s |
| Released | September 2, 2025 | April 24, 2026 |
| Parameters | — | 284B (13B active) |
| Input modalities | — | Text |
| Output modalities | — | Text |
Shared benchmarks
Apertus 70B Instruct
DeepSeek-V4-Flash
GPQA Diamond
27.2
89.4
Humanity’s Last Exam
5.5
32.1