Phi-4
By Microsoft · released 2024-12-12 (22 months) · snapshot 2026-10-24
- Intelligence index
- 5.90
- Price per 1M tokens (in / out)
- $0.13 / $0.50
- End-to-end time
- 15.5s
Is Phi-4 on the Pareto frontier, and what beats it?
Not ranked · benchmark data missing
Phi-4 does not report both selected axis metrics.
Open in the picker Daily check URL: 200 on the frontier, 409 once beaten
Phi-4 today
Incomplete benchmark data
Intelligence and benchmark cost are both required to place this model on the frontier. Missing data is not a performance verdict.
- Intelligence index
- 5.90
- Benchmark cost
- Not reported
- Input / 1M tokens
- 0.13
- Output / 1M tokens
- 0.50
- Context window
- 16k
Intelligence: Artificial Analysis Intelligence Index. Cost: total cost to run that benchmark, so reasoning tokens count. API id: phi-4.
Compare head to head
Should you use Phi-4?
Should you use Phi-4?
No verdict: it does not report both intelligence and benchmark cost.
What is better and cheaper than Phi-4?
Other models on the intelligence-vs-cost frontier
How much does Phi-4 cost?
$0.13/$0.50 in/out per 1M tokens, and - to run the full Artificial Analysis Intelligence Index, the honest all-in number for reasoning models.
Frontier membership by axis pair
Other axis pairs where Phi-4 is Pareto-optimal.
| Y axis | X axis | On the frontier |
|---|---|---|
| AA intelligence index | AA total cost to run (USD) | no data |
| AA intelligence index | End-to-end total seconds | |
| AA intelligence index | Output speed tokens/sec |
Accept: text/markdown or ?format=md returns this page as Markdown. JSON: /api/v1/models/phi-4 · Markdown: /api/v1/models/phi-4?format=md · llms.txt · agent skill