DeepSeek-V4-Flash vs Llama Guard 3 8B
DeepSeek vs Meta — benchmarks, pricing, and capabilities side by side.
- •DeepSeek-V4-Flash is cheaper ($0.10 vs $0.48 per 1M input)
- •DeepSeek-V4-Flash has a larger context window (1M)
| DeepSeek-V4-Flash | Llama Guard 3 8B | |
|---|---|---|
| Intelligence index | 89.4 | — |
| Developer | DeepSeek | Meta |
| Type | LLM | LLM |
| Access | Open weights | Open weights |
| Context window | 1,048,576 tokens | 131,072 tokens |
| Input price | $0.10 / 1M | $0.48 / 1M |
| Output price | $0.20 / 1M | $0.03 / 1M |
| Speed | 109 tok/s | — |
| Released | April 24, 2026 | February 12, 2025 |
| Parameters | 284B (13B active) | — |
| Input modalities | Text | Text |
| Output modalities | Text | Text |