DeepSeek V4 Flash vs Auto model (Standard)
DeepSeek V4 Flash is cheaper on output tokens, while DeepSeek V4 Flash offers a larger context window. Choose DeepSeek V4 Flash or Auto model (Standard) based on the trade-off between cost, context, and the benchmarks that matter for your use case.
| Spec | DeepSeek V4 Flash | Auto model (Standard) |
|---|---|---|
| Provider | NanoGPT | NanoGPT |
| Input / 1M tokens | $0.14 | $10.00 |
| Output / 1M tokens | $0.28 | $19.99 |
| Context window | 1M | 1M |
| Parameters | 284B | — |
| Open weights | No | No |
| Released | Apr 2026 | Jun 2024 |
FAQ
Pricing is indicative — confirm with the provider before production use. Highlighted values indicate the better figure for that row.