Key dates, capabilities, and model metadata.
30 Apr 2025
Parameters
3,800,000,000
License
MIT
Training Tokens
150,000,000,000
Input
No modalities listed.
Output
Headline benchmark standings and comparison context.
Top benchmark results for microsoft/phi-4-mini-reasoning.
Detailed benchmark comparisons now live in the Compare tool.
No modalities listed.
arxiv.org
No API pricing or subscription plan information is available for this model yet.
If you know providers we can integrate, please tell us on Discord or open an issue on GitHub so we can add pricing data.Open an issue
