All models
Microsoft
Phi 4 fastruns locally
microsoft/phi-4
[Microsoft Research](/microsoft) Phi-4 is designed to perform well in complex reasoning tasks and can operate efficiently in situations with limited memory or where quick responses are needed. At 14 billion...
Input / 1M
$0.070
USD per million tokens
Output / 1M
$0.14
USD per million tokens
Est. cost / token
$0.000000087
$0.088 per 1M, 3:1 mix
Context window
16K
max output 15K
Major features
Open weightsJSON schema
Task suitability (heuristic 1–5, not a benchmark)
- coding3
- reasoning2
- math2
- summarization4
- extraction4
- classification4
- chat3
- creative3
- translation4
- vision1
- long context1
- agentic1
Run it locally
Will it run on my machine? →ollama run phi4:14b| Quantization | Quality | Memory needed (8K ctx) |
|---|---|---|
| Q2_K | poor | 8.4 GB |
| Q3_K_M | fair | 9.5 GB |
| Q4_K_Mrecommended | good | 11.4 GB |
| Q5_K_M | very good | 13 GB |
| Q6_K | excellent | 14.6 GB |
| Q8_0 | near-lossless | 18.4 GB |
| F16 | lossless | 32.9 GB |
Estimates.
Price history
| Recorded | Input | Output | Cached | Source |
|---|---|---|---|---|
| 2026-11-05 03:20 | $0.070 | $0.14 | — | openrouter |
A row is added whenever a sync detects a price change. Vendor pricing source
Pricing (USD per 1M tokens)
- Input
- $0.070
- Output
- $0.14
- Blended (3:1)
- $0.088
Details
- Company
- Microsoft
- Version
- 4
- Family
- Phi
- Released
- 2025-01-10
- Parameters
- 14.7B
- Max output
- 15K
- Open weights
- Yes
- Inputs
- text
- Last synced
- 46 min ago