Model
Model
deepseek
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 | Quantization | Input / M tokens | Output / M tokens |
|---|---|---|---|
| Relacerelace/fp4 | fp4 | $0.05 | $0.45 |
| StreamLakestreamlake/fp8 | fp8 | $0.0826 | $0.1652 |
| DeepInfradeepinfra/fp8 | fp8 | $0.09 | $0.18 |
| GMICloudgmicloud/fp8 | fp8 | $0.091 | $0.182 |
| Venicevenice | — | $0.0966 | $0.1925 |
| DigitalOceandigitalocean | — | $0.098 | $0.196 |
| SiliconFlowsiliconflow/fp8 | fp8 | $0.13 | $0.28 |
| AtlasCloudatlas-cloud/fp4 | fp4 | $0.14 | $0.28 |
| Baidubaidu/fp8 | fp8 | $0.14 | $0.28 |
| Parasailparasail/fp8 | fp8 | $0.14 | $0.28 |
| Alibabaalibaba/fp8/us-east-1 | fp8 | $0.15 | $0.30 |
| Mancer 2mancer/fp8 | fp8 | $0.19 | $0.50 |
| Azureazure/us | — | $0.21 | $0.56 |
| Cloudflarecloudflare | — | $0.44 | $1.32 |
| Waferwafer/mxfp4 | mxfp4 | $0.70 | $0.17 |
Prices are what the Provider charges, synced 2026-10-21. Amber adds its platform fee to each Trial’s total; see pricing.
deepseek-v4-flash
deepseek-v4-flash-high