MicrosoftReleased on Apr 30, 2025

Phi 4 Mini Reasoning: API Pricing, Context Window & Benchmarks

Phi 4 Mini Reasoning is a language model from Microsoft, released in April 2025.

Phi-4-mini-reasoning is designed for multi-step, logic-intensive mathematical problem-solving tasks under memory/compute constrained environments and latency bound scenarios. Some of the use cases include formal proof generation, symbolic

Phi 4 Mini Reasoning benchmarks

Rankings

Quality Tracker

Phi 4 Mini Reasoning Performance Across Datasets

Scores sourced from the model's scorecard, paper, or official blog posts

LLM Stats Logollm-stats.com - Wed Jul 29 2026
Notice missing or incorrect data?

Phi 4 Mini Reasoning model size

Phi 4 Mini Reasoning has 3.8 billion parameters and was trained on 150 billion tokens. See how it compares to other models in the same parameter range.

ParametersTraining tokens
3.8B
150Btokens
39× tokens-to-params ratio
Small (3–10B)
3.8B
1B7B70B405B

Phi 4 Mini Reasoning API

Available from the model provider

Phi 4 Mini Reasoning has an official provider API. It is not currently routed through the LLM Stats gateway.

Read the official API documentation

Phi 4 Mini Reasoning latency

Phi 4 Mini Reasoning time to first token, sustained output throughput, and failed-request rate from live API traffic over the trailing 7 days.

Phi 4 Mini Reasoning examples

Recent arena outputs from Phi 4 Mini Reasoning, picked from the highest-ranked matchups.

Phi 4 Mini Reasoning license

Phi 4 Mini Reasoning is released under the MIT license, which permits commercial use, has 3.8B parameters, has a knowledge cutoff of February 2025.

License
MIT
Commercial use allowed
Parameters
3.8B
Knowledge cutoff
February 2025

MIT License - allows commercial use

Phi 4 Mini Reasoning resources

Official sources for Phi 4 Mini Reasoning: api documentation, paper or system card, official launch post, model weights.

Phi 4 Mini Reasoning vs other models

The most-compared alternatives to Phi 4 Mini Reasoning are MiMo-V2.5-Pro, GPT-4o, DeepSeek R1 Zero. Open any pair side-by-side for benchmarks, pricing, context, and latency.

Models like Phi 4 Mini Reasoning

Models ranked just above and below Phi 4 Mini Reasoning by LLM Stats score.

 

MiMo-V2.5-Pro

Score pending
DetailsCompare
 

GPT-4o

Score pending
DetailsCompare
 

DeepSeek R1 Zero

Score pending
DetailsCompare
 

DeepSeek R1 Distill Llama 70B

Score pending
DetailsCompare
 

DeepSeek R1 Distill Qwen 32B

Score pending
DetailsCompare
 

Min istral 3 (3B Reasoning 2512)

Score pending
DetailsCompare

FAQ

Common questions about Phi 4 Mini Reasoning.

When was Phi 4 Mini Reasoning released?

Phi 4 Mini Reasoning was released on April 30, 2025 by Microsoft. This is the official Phi 4 Mini Reasoning release date tracked on LLM Stats.

Is Phi 4 Mini Reasoning available via API?

Yes, Phi 4 Mini Reasoning is available via API. See the official documentation for authentication and endpoint details.

How big is Phi 4 Mini Reasoning?

Phi 4 Mini Reasoning has 3.8 billion parameters. It was trained on 150 billion tokens. It ships as an open-weight model, so you can download and run it on your own hardware.

Who created Phi 4 Mini Reasoning?

Phi 4 Mini Reasoning was created by Microsoft.

What is the license for Phi 4 Mini Reasoning?

Phi 4 Mini Reasoning is released under the MIT license. This is an open-source / open-weight license that permits self-hosting.

What is the knowledge cutoff date for Phi 4 Mini Reasoning?

Phi 4 Mini Reasoning has a knowledge cutoff of February 2025, meaning it was trained on data up to that point and may not know about events after it.

Where is the Phi 4 Mini Reasoning paper or technical report?

Phi 4 Mini Reasoning has a paper or technical report available at https://arxiv.org/pdf/2504.21233. Use that source for architecture, training, release and evaluation details.

What models should I compare Phi 4 Mini Reasoning against?

Common Phi 4 Mini Reasoning comparisons include Phi 4 Mini Reasoning vs MiMo-V2.5-Pro, Phi 4 Mini Reasoning vs GPT-4o, Phi 4 Mini Reasoning vs DeepSeek R1 Zero. Compare them side by side for benchmark scores, pricing, context window, latency and API availability.