| per 1M tokens | DeepSeekDeepSeek-V4 Flash | OpenAIGPT-5.4 nano |
|---|---|---|
| Input | $0.14 | $0.20 |
| Output | $0.28 | $1.25 |
| Cache Read | $0.0028 | $0.02 |
| Batch Input | — | $0.10 |
| Batch Output | — | $0.625 |
| Est / Month* | $12.60 | $39.25 |
| API Model ID | deepseek-v4-flash | gpt-5.4-nano |
| Context Window | 1M tokens | 400K tokens |
| Max Output | 384K tokens | 128K tokens |
| modalities | text | text · image |
| capabilities | tool-use · streaming · json-mode · code · reasoning | tool-use · streaming · json-mode · vision · code |
Text / Chat · cheapest per row in red
* At the reference workload: 2,000 input / 1,250 output tokens per request, 20,000 requests per month. Excludes cache writes and batch discounts.