DeepSeek V4 Flash is an efficiency-optimized Mixture-of-Experts model from DeepSeek with 284B total parameters and 13B activated parameters, supporting a 1M-token context window. It is designed for fast inference and...
| Provider | Input $/M | Output $/M | Blended $/M ↑ | vs Median |
|---|---|---|---|---|
| ↓AtlasCloud | $0.08 | $0.17 | $0.13 | -40.0% |
| StreamLake | $0.09 | $0.18 | $0.14 | -35.6% |
| DeepInfra | $0.09 | $0.18 | $0.14 | -34.6% |
| GMICloud | $0.09 | $0.18 | $0.15 | -33.9% |
| DigitalOcean | $0.10 | $0.20 | $0.16 | -28.8% |
| Alibaba | $0.13 | $0.27 | $0.21 | -2.6% |
| SiliconFlow | $0.13 | $0.28 | $0.22 | -0.1% |
| Venice | $0.14 | $0.28 | $0.22 | median |
| Baidu | $0.14 | $0.28 | $0.22 | +1.7% |
| Novita | $0.14 | $0.28 | $0.22 | +1.7% |
| Parasail | $0.14 | $0.28 | $0.22 | +1.7% |
| OpenInference | $0.04 | $0.50 | $0.32 | +43.5% |
| Mancer 2 | $0.19 | $0.50 | $0.38 | +70.8% |
| Azure | $0.21 | $0.56 | $0.42 | +90.7% |
| Relace | $0.03 | $1.28 | $0.78 | +254.2% |