phi-4 API Pricing
phi-4 on DeepInfra costs $0.07 per 1M input tokens and $0.14 per 1M output tokens (as of ).
| Input | $0.07 / 1M tokens |
|---|---|
| Output | $0.14 / 1M tokens |
| Cached input (read) | — |
| Cache write | — |
| Batch input / output | — |
| Above 200K tokens (input / output) | — |
| Context window | 16K tokens |
| Max output | 16K tokens |
| Supports | tool calling |
| Deprecation date | None listed |
| API model id | microsoft/phi-4 |
| In our records since |
What it costs for common jobs
Estimated from the list price above (no caching or batch discount).
| Job | Estimated cost |
|---|---|
| 1,000 chat replies (2,000 tokens in, 500 out each) | $0.21 |
| Summarising 100 long documents (50,000 in, 1,000 out each) | $0.36 |
| 1 million classification calls (300 in, 5 out each) | $21.70 |
Try your own numbers in the LLM API cost calculator.
Price history
No price change recorded since .
Similarly priced models
| Model | Provider | Input / 1M | Output / 1M | Cached input / 1M | Context |
|---|---|---|---|---|---|
| Nemotron-3-Nano-30B-A3B | DeepInfra | $0.05 | $0.20 | $0.03 | 262K |
| Ling-3.0-flash | DeepInfra | $0.06 | $0.18 | $0.01 | 131K |
| gemma-3-12b-it | DeepInfra | $0.05 | $0.15 | — | 131K |
| gemma-3-27b-it | DeepInfra | $0.08 | $0.16 | — | 131K |
| gpt-oss-120b | DeepInfra | $0.04 | $0.17 | — | 131K |
| DeepSeek-V4-Flash-0731 | DeepInfra | $0.08 | $0.18 | $0.02 | 1.04858M |
| NVIDIA-Nemotron-Nano-9B-v2 | DeepInfra | $0.04 | $0.16 | — | 131K |
| Mistral-Small-3.2-24B-Instruct-2506 | DeepInfra | $0.07 | $0.20 | — | 128K |