Microsoft: Phi 4
The Microsoft Phi 4 is a text-based AI model that processes input up to 16,384 tokens and generates output with a maximum of 14,745 completion tokens. It only accepts text input modality. For users looking for a text-only AI solution, the Phi 4 may be worth considering due to its competitive pricing: $0.07 per million input tokens and $0.14 per million output tokens. Its blended benchmark score of 4.4 across one independent benchmark suggests it can perform at an above-average level in certain tasks. However, users with complex requirements that involve tools, reasoning, or structured outputs may want to explore other options.
Benchmark results
Independent, published benchmarks. Blended score 4.0 across 1 benchmark, last refreshed 2026-10-13. How scoring works →
| Benchmark | Measures | Score |
|---|---|---|
| AI Index | broad capability composite | 9.8 |
- Model ID
- microsoft/phi-4
- Vendor
- microsoft
- Released
- January 2025
- Tokenizer
- Other
- Input Modalities
- text
- Output Modalities
- text
- Max Output
- 14,745 tokens
- Tool Calling
- not supported
- Structured Output
- ✓ supported
- Reasoning Mode
- not supported
- Vision
- text only
- Audio
- no
- Moderated
- no
What it costs in practice
Computed from the current $0.07/M input and $0.14/M output rates. Run your own numbers →
| Job | Tokens | Cost |
|---|---|---|
| Summarize a 50-page report | 30k in / 1.5k out | under $0.01 |
| Classify 1,000 customer emails | 500k in / 50k out | $0.04 |
| A month of a busy support chatbot | 5M in / 2M out | $0.63 |
Similar models
Quick answers
- How much does Microsoft: Phi 4 cost?
- $0.07 per million input tokens and $0.14 per million output tokens.
- What is Microsoft: Phi 4's context window?
- 16,384 tokens, roughly 24 pages of text in a single request.
- Does Microsoft: Phi 4 support tool calling?
- It supports structured output.
- Can Microsoft: Phi 4 process images?
- No, it is text-only on the input side.