DeepSeek-V4-Flash vs Qwen3 Omni 30B A3B Instruct
DeepSeek vs Alibaba — benchmarks, pricing, and capabilities side by side.
- •DeepSeek-V4-Flash has the higher intelligence index (89.4 vs 57.3)
- •DeepSeek-V4-Flash is cheaper ($0.10 vs $0.30 per 1M input)
- •DeepSeek-V4-Flash is faster
| DeepSeek-V4-Flash | Qwen3 Omni 30B A3B Instruct | |
|---|---|---|
| Intelligence index | 89.4 | 57.3 |
| Developer | DeepSeek | Alibaba |
| Type | LLM | LLM |
| Access | Open weights | — |
| Context window | 1,048,576 tokens | — |
| Input price | $0.10 / 1M | $0.30 / 1M |
| Output price | $0.20 / 1M | $1.00 / 1M |
| Speed | 109 tok/s | 103 tok/s |
| Released | April 24, 2026 | September 22, 2025 |
| Parameters | 284B (13B active) | — |
| Input modalities | Text | — |
| Output modalities | Text | — |
Shared benchmarks
DeepSeek-V4-Flash
Qwen3 Omni 30B A3B Instruct
GPQA Diamond
89.4
62
Humanity’s Last Exam
32.1
5.1