Emerging model from Microsoft. Released Apr 2024. Benchmark data.
📊
Benchmark-Only Model
This model appears in the Epoch AI Capabilities Index but lacks full data coverage. Benchmark scores are available below, but pricing, compatibility, and other detailed metrics may be limited.
Metrics
Performance
Context Window0
Maximum input token capacity
Info
ProviderMicrosoft
Organization that created the model
CountryUSA
Model development country
Open WeightsNo
Model weights publicly available
Open SourceNo
Open source availability
Epoch Model IDPhi-3-mini-4k-instruct
Epoch AI benchmark identifier
Score Breakdown
code
20%70
Code generation, understanding, and debugging
math
20%70
Mathematical reasoning and problem solving
reasoning
15%70
Multi-step logical reasoning
intelligence
30%70
Overall reasoning and task completion ability
instruction following
70
Compatibility
No compatibility data
Sources(1)
Benchmark Performance
Compare Models| Dimension | Score |
|---|---|
| Intelligence | 70.0 |
| Reasoning | 70.0 |
| Math | 70.0 |
| Code | 70.0 |
| Instruction Following | 70.0 |
| Overall | 70 |
Loading chart...
Data: Epoch AI (CC-BY 4.0)