The AI arena is free today

Open Superagent

Gemini Diffusion vs Phi 4

Gemini Diffusion and Phi 4 are closely matched at 0.9 and 5.4 on the LLM Stats Score.

Google · Microsoft · Updated for 2026

Which is better?

Gemini Diffusion and Phi 4 are closely matched on the overall LLM Stats Score at 0.9 and 5.4.

The models split the 2 individual benchmarks reported for both models evenly.

Based on current LLM Stats indexes, shared benchmarks, pricing, and model metadata for 2026.

Choose Gemini Diffusion

  • you want the most recent training data — it shipped May 2025

Choose Phi 4

  • you need open weights you can self-host or fine-tune

At a glance

The differences that matter most.

Gemini Diffusion
Google
Phi 4
Microsoft
Core performance indexes
LLM Stats Score
0.9
#329
5.4
#302
Reasoning
0.9
#321
6.5
#293
Coding
2.8
#227
3.4
#223
Cost, coverage & limits
Benchmark wins
1 of 2
1 of 2
Input price
— / M
$0.07 / M
Output price
— / M
$0.14 / M
Context window
—
16,384

Individual benchmarks

10 reported for Gemini Diffusion · 13 for Phi 4

2 shared

Gemini Diffusion outperforms in 1 benchmarks (HumanEval), while Phi 4 is better at 1 benchmark (GPQA).

Both models are evenly matched across the benchmarks.

Wed Sep 23 2026 • llm-stats.com

Human preference

Blind head-to-head votes and playground preference scores

Context Window

Maximum input and output token capacity

Only Phi 4 specifies input context (16,384 tokens). Only Phi 4 specifies output context (16,384 tokens).

Google
Gemini Diffusion
Input- tokens
Output- tokens
Microsoft
Phi 4
Input16,384 tokens
Output16,384 tokens
Wed Sep 23 2026 • llm-stats.com

License

Usage and distribution terms

Gemini Diffusion is licensed under a proprietary license, while Phi 4 uses MIT.

License differences may affect how you can use these models in commercial or open-source projects.

Gemini Diffusion

Proprietary

Closed source

Phi 4

MIT

Open weights

Release Timeline

When each model was launched

Gemini Diffusion was released on 2025-05-20, while Phi 4 was released on 2024-12-12.

Gemini Diffusion is 5 months newer than Phi 4.

Gemini Diffusion

May 20, 2025

1.3 years ago

5mo newer
Phi 4

Dec 12, 2024

1.8 years ago

Knowledge Cutoff

When training data ends

Phi 4 has a documented knowledge cutoff of 2024-06-01, while Gemini Diffusion's cutoff date is not specified.

We can confirm Phi 4's training data extends to 2024-06-01, but cannot make a direct comparison without Gemini Diffusion's cutoff date.

Gemini Diffusion

—

Phi 4

Jun 2024

Outputs Comparison

Notice missing or incorrect data?Start an Issue discussion→

Judge for yourself.

Run your own prompts against Gemini Diffusion and Phi 4 side-by-side, then vote on the output you prefer.

Gemini Diffusion
✓ Preferred
Phi 4
Open in Playground

FAQ

Common questions about Gemini Diffusion vs Phi 4.

Which is better, Gemini Diffusion or Phi 4?

Gemini Diffusion and Phi 4 are closely matched on the LLM Stats Score at 0.9 and 5.4. Gemini Diffusion is made by Google and Phi 4 is made by Microsoft. The best choice depends on your use case — compare their capability indexes, individual benchmarks, pricing, and limits above.

How does Gemini Diffusion compare to Phi 4 in benchmarks?

Gemini Diffusion scores HumanEval: 89.6%, MBPP: 76.0%, Global-MMLU-Lite: 69.1%, LBPP (v2): 56.8%, BigCodeBench: 45.4%. Phi 4 scores MMLU: 84.8%, HumanEval+: 82.8%, HumanEval: 82.6%, MGSM: 80.6%, MATH: 80.4%.

What are the context window sizes for Gemini Diffusion and Phi 4?

Gemini Diffusion supports an unknown number of tokens and Phi 4 supports 16K tokens. A larger context window lets you process longer documents, conversations, or codebases in a single request.

What are the main differences between Gemini Diffusion and Phi 4?

Key differences include LLM Stats Score (0.9 vs 5.4), licensing (Proprietary vs MIT). See the full comparison above for benchmark-by-benchmark results.

Who makes Gemini Diffusion and Phi 4?

Gemini Diffusion is developed by Google and Phi 4 is developed by Microsoft.

Related comparisons