| Nova Lite 1.0 | DeepSeek V4 Flash 0731 | |
|---|---|---|
| Input, per 1M tokens | $0.06 | $0.021 |
| Output, per 1M tokens | $0.24 | $0.32 |
| Context window | 300K | 1.31M |
| Classification or routing (600 tokens in, 10 out), per 1,000 | $0.0384 | $0.0158 |
| Chat reply (2,000 in, 400 out), per 1,000 | $0.216 | $0.17 |
| Retrieval-augmented answer (8,000 in, 500 out), per 1,000 | $0.60 | $0.328 |
| Confidence signal | no | yes |
| Tool calling | yes | yes |
| BANKING77, correct of 400 | not measured | 298 |
Which one should carry your traffic?
If DeepSeek V4 Flash 0731 answers your requests as well as Nova Lite 1.0, moving that traffic saves 21% on chat-shaped requests. Averages cannot tell you whether it does for your requests. DeepSeek V4 Flash 0731 exposes a confidence signal, so Gimbix can let it answer first and hand unsure cases to Nova Lite 1.0.
Nova Lite 1.0 profile · DeepSeek V4 Flash 0731 profile · Measure it on your traffic