DeepSeek V4 Flash 0423 is 51.3× cheaper per million tokens (blended). DeepSeek V4 Flash 0423 has the larger context window (1,048,575 vs 65,536 tokens, 16.0×). DeepSeek V4 Flash 0423 is open-weights; the other is not.
DeepSeek V4 Flash 0423 vs Nano Banana Pro (Gemini 3 Pro Image Preview)
Live catalog fields. Best-in-row is highlighted.
DeepSeek V4 Flash 0423open
DeepSeek · deepseek__deepseek-v4-flash
Nano Banana Pro (Gemini 3 Pro Image Preview)
Google · google__gemini-3-pro-image-preview
Identity
| Field | deepseek__deepseek-v4-flash | google__gemini-3-pro-image-preview |
|---|---|---|
| Provider | DeepSeek | |
| Slug | deepseek__deepseek-v4-flash | google__gemini-3-pro-image-preview |
| Status | live | live |
| Open weights | Yes | No |
| License | — | — |
| HuggingFace | deepseek-ai/DeepSeek-V4-Flash | — |
Quality
| Field | deepseek__deepseek-v4-flash | google__gemini-3-pro-image-preview |
|---|---|---|
| Quality index (0–100) | — | — |
Cost
| Field | deepseek__deepseek-v4-flash | google__gemini-3-pro-image-preview |
|---|---|---|
| Input $ / 1M tokens | $0.09 | $2.00 |
| Cached input $ / 1M | $0.02 | $0.20 |
| Output $ / 1M tokens | $0.18 | $12.00 |
Context
| Field | deepseek__deepseek-v4-flash | google__gemini-3-pro-image-preview |
|---|---|---|
| Max context tokens | 1,048,575 | 65,536 |
| Max output tokens | 943,717 | 32,768 |
Modalities
| Field | deepseek__deepseek-v4-flash | google__gemini-3-pro-image-preview |
|---|---|---|
| Input | text | image, text |
| Output | text | image, text |