No longer supported
This model is legacy (released June 2023) and is no longer actively maintained or recommended by Microsoft. Consider using their current flagship model instead.
by Microsoft· Released June 2023
Phi-1 is a small language model (1.3B parameters) developed by Microsoft Research, trained on a curated dataset of textbook-quality code and synthetic exercises. It achieves state-of-the-art performance on Python coding benchmarks among models under 10B parameters, demonstrating that high-quality data can compensate for scale. Phi-1 is part of Microsoft's Phi series, which includes Phi-1.5 and Phi-2.
Input cost
Free (open source)
Output cost
Free (open source)
Context window
2048 tokens
Max output
—
Modalities
Parameters
1.3B
License
MIT
Python code generation and understanding tasks, especially for educational and research purposes.