Phi-4 is a text-based AI model by Microsoft Azure. It features 12 tok/s output speed, $0.13/1M input tokens pricing. It scores 10.4 on the intelligence index.
Output Speed
12 tok/s
Latency (TTFT)
0.61s
Blended Price
$0.22/M
| Input (Prompt) | $0.13 |
| Output (Completion) | $0.50 |
| Cache Read | Free |
| Cache Write | Free |
| Context Length | N/A |
| Max Output Tokens | N/A |
| Input Modalities | Text |
| Output Modalities | Text |
| Tokenizer | N/A |
Phi-4 costs $0.13/1M input tokens and $0.50/1M output tokens.
Phi-4 has a lower coding score of 11.2. For demanding coding tasks, consider a model with a higher coding benchmark.
Phi-4 generates output at 12 tok/s. Time to first token is 0.61s.
No, Phi-4 is a paid model. Check the free models page for zero-cost alternatives.
See the alternatives section above for models with similar capabilities. You can also compare Phi-4 head-to-head with any model on our comparison page.
Last updated: