| Phi 4 | GPT-5.4 Nano | |
|---|---|---|
| Input, per 1M tokens | $0.07 | $0.20 |
| Output, per 1M tokens | $0.14 | $1.25 |
| Context window | 16K | 400K |
| Classification or routing (600 tokens in, 10 out), per 1,000 | $0.0434 | $0.133 |
| Chat reply (2,000 in, 400 out), per 1,000 | $0.196 | $0.90 |
| Retrieval-augmented answer (8,000 in, 500 out), per 1,000 | $0.63 | $2.23 |
| Confidence signal | no | no |
| Tool calling | no | yes |
Which one should carry your traffic?
If Phi 4 answers your requests as well as GPT-5.4 Nano, moving that traffic saves 78% on chat-shaped requests. Averages cannot tell you whether it does for your requests. Phi 4 does not expose a confidence signal through OpenRouter, so Gimbix would test it as a full replacement rather than a first answer.
Phi 4 profile · GPT-5.4 Nano profile · Measure it on your traffic