← Model list
Phi-4-mini
microsoft·microsoft/phi-4-mini·GA·Open weights·phi series
Compact Microsoft instruction model tuned for efficient coding assistance, reasoning, and low-latency agent tasks
At a glance
Official price$0.075 / $0.30 per 1M
Cache read — · Azure Cognitive Services
Lowest paid$0.075 / $0.30
Azure Cognitive Services First-party · 2.3× spread
1 more $0 channels
Context128,000
⚠ Providers report 128,000–131,072; the table below is authoritative
Output limit4,096
Capabilities
Reasoning✓ Tool useStructured output✓ TemperatureAttachments
⚠ Providers report capability flags inconsistently
Modalities
Text
Knowledge cutoff2023-10
Released / updated2024-12-11 / 2024-12-11
Quality & performanceArtificial Analysis · Intelligence Index v4.1
Intelligence5.7
Value43
Coding3.8
Agentic—
Output speed44 tok/s
TTFT0.86 s
Value formulaIQ 5.7 ÷ min blended $0.131 = 43
Quality is independently evaluated by Artificial Analysis. Speed/latency are model-level medians.
Available at 4 providers3 with public prices · 1 free
| Provider | Tier | Input | Output | Cache read | Cache write | Context | Output limit | Status |
|---|---|---|---|---|---|---|---|---|
| Nvidia microsoft/phi-4-mini-instruct | First-party | Free | — | — | 131,072 ⚠ | 8,192 | ||
| Azure Cognitive ServicesOfficial | First-party | $0.075 | $0.30 | — | — | 128,000 | 4,096 | |
| AzureOfficial | First-party | $0.075 | $0.30 | — | — | 128,000 | 4,096 | |
| NanoGPT phi-4-mini-instruct | Gateway | $0.17 | $0.68 | $0.085 | — | 128,000 | 16,384 |
Sorted by blended price (input×0.75 + output×0.25) asc. The official channel always shows regardless of rank. Whether a gateway's low price is actually usable can't be verified.
Your usage cost
1Azure Cognitive Services · Official$30.00
2Azure · Official$30.00
3NanoGPT$57.80
1 more channels offer $0 (Nvidia); free tiers usually have rate limits and no SLA, excluded from ranking.
The cheapest paid channel is official.
Benchmark1 items
| Name | Conditions | Score | Metric | Source |
|---|---|---|---|---|
| MMLU | — | 67.3 | accuracy | Source ↗ |
The same benchmark scores very differently across harness / dataset, so the qualifying conditions must be shown together.
Related models
Phi 4 Multimodalsame series$0.08 / $0.32Phi-4same series$0.125 / $0.50Microsoft: Phi 4same series$0.07 / $0.14Llama 4 Maverick 17B 128E Instruct FP8cheaper alternative$0 / $0Lyria 3 Clip Previewcheaper alternative$0 / $0Command R7Bcheaper alternative$0.037 / $0.15Granite 4.1 8Bcheaper alternative$0.05 / $0.10
Price historyone sample accumulated per data sync
Input listOutput listMin blended