| Gemini 3.8 Flash | Qwen3.8 27B | |
|---|---|---|
| Input, per 1M tokens | $0.75 | $0.42 |
| Output, per 1M tokens | $3.75 | $3.00 |
| Context window | 1.05M | 1M |
| Classification or routing (600 tokens in, 10 out), per 1,000 | $0.487 | $0.282 |
| Chat reply (2,000 in, 400 out), per 1,000 | $3.00 | $2.04 |
| Retrieval-augmented answer (8,000 in, 500 out), per 1,000 | $7.88 | $4.86 |
| Confidence signal | no | yes |
| Tool calling | yes | yes |
| BANKING77, correct of 400 | 331 | 264 |
Which one should carry your traffic?
If Qwen3.8 27B answers your requests as well as Gemini 3.8 Flash, moving that traffic saves 32% on chat-shaped requests. Averages cannot tell you whether it does for your requests. Qwen3.8 27B exposes a confidence signal, so Gimbix can let it answer first and hand unsure cases to Gemini 3.8 Flash.
Gemini 3.8 Flash profile · Qwen3.8 27B profile · Measure it on your traffic