DeepSeek V4 Flash 0731 is 23.0× cheaper per million tokens (blended). DeepSeek V4 Flash 0731 has the larger context window (1,048,576 vs 262,144 tokens, 4.0×).
DeepSeek V4 Flash 0731 vs Qwen3.8 27B
Live catalog fields. Best-in-row is highlighted.
DeepSeek V4 Flash 0731open
DeepSeek · deepseek__deepseek-v4-flash-0731
Qwen3.8 27Bopen
Qwen · qwen__qwen3.8-27b
Identity
| Field | deepseek__deepseek-v4-flash-0731 | qwen__qwen3.8-27b |
|---|---|---|
| Provider | DeepSeek | Qwen |
| Slug | deepseek__deepseek-v4-flash-0731 | qwen__qwen3.8-27b |
| Status | live | live |
| Open weights | Yes | Yes |
| License | — | — |
| HuggingFace | deepseek-ai/DeepSeek-V4-Flash-0731 | Qwen/Qwen3.8-27B |
Quality
| Field | deepseek__deepseek-v4-flash-0731 | qwen__qwen3.8-27b |
|---|---|---|
| Quality index (0–100) | — | — |
Cost
| Field | deepseek__deepseek-v4-flash-0731 | qwen__qwen3.8-27b |
|---|---|---|
| Input $ / 1M tokens | $0.04 | $0.21 |
| Cached input $ / 1M | $0.02 | $0.15 |
| Output $ / 1M tokens | $0.08 | $2.55 |
Context
| Field | deepseek__deepseek-v4-flash-0731 | qwen__qwen3.8-27b |
|---|---|---|
| Max context tokens | 1,048,576 | 262,144 |
| Max output tokens | 943,718 | 131,072 |
Modalities
| Field | deepseek__deepseek-v4-flash-0731 | qwen__qwen3.8-27b |
|---|---|---|
| Input | text | text, image, video |
| Output | text | text |