Nova 2 Lite vs DeepSeek V4 Flash
Amazon against DeepSeek, costed on the same three workloads. Neither is cheaper everywhere — which one wins depends on how much you read versus write.
Monthly bill, side by side
| Workload | Nova 2 Lite | DeepSeek V4 Flash | Difference |
|---|---|---|---|
| Support chatbot 2,500 six-turn conversations a day |
$1,613 | $1,416 | $198 (+14%) |
| Document summariser 1,000 single-shot summaries a day |
$170 | $193 | $22.92 (+13%) |
| Coding agent 200 twelve-turn sessions a day |
$1,022 | $1,077 | $54.72 (+5%) |
No prompt caching and no batch discount on either side, so the comparison is like for like. Both figures assume 30 days a month.
The rate cards
| Nova 2 Lite | DeepSeek V4 Flash | |
|---|---|---|
| Input, per 1M tokens | $0.3 | $0.44 |
| Output, per 1M tokens | $2.5 | $1.32 |
| Cached input | $0.075 | $0.014 |
| Context window | 1M | 1M |
| Maximum output | 64K | 384K |
| Reasoning model | yes | yes |
| Vision | yes | no |
Run this comparison on your own numbers
Full detail: Nova 2 Lite pricing · DeepSeek V4 Flash pricing