The AI arena is free today

Open Superagent

IBM Granite 4.0 Tiny Preview vs Phi 4

Phi 4 leads the LLM Stats Score 5.4 to -5.5.

IBM · Microsoft · Updated for 2026

Which is better?

Phi 4 leads the overall LLM Stats Score 5.4 to -5.5, ranking #296 overall.

In the 6 individual benchmarks reported for both models, Phi 4 wins 6; this is a narrower head-to-head signal than the composite indexes.

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

Choose IBM Granite 4.0 Tiny Preview

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

Choose Phi 4

  • overall performance matters — it scores 5.4 and ranks #296 on LLM Stats
  • your work emphasizes reasoning — it leads those capability indexes
  • you value its reported benchmark strengths — it wins 6 of 6 exact shared results

At a glance

The differences that matter most.

Core performance indexes
-5.5
#358
5.4
#296
-5.5
#348
6.5
#287
3.0
#220
3.4
#217
Cost, coverage & limits
Benchmark wins
0 of 6
6 of 6
Input price
— / M
$0.07 / M
Output price
— / M
$0.14 / M
Context window
16,384

Capability indexes

Additional strengths measured across groups of related public benchmarks

3 shared
Index
IBM Granite 4.0 Tiny Preview
Phi 4
-5.8#318
12.8#230
-3.7#208
12.9#137
-3.7#193
12.9#120
Conservative TrueSkill rating · higher is betterHow scores work

Individual benchmarks

12 reported for IBM Granite 4.0 Tiny Preview · 13 for Phi 4

6 shared

IBM Granite 4.0 Tiny Preview outperforms in 0 benchmarks, while Phi 4 is better at 5 benchmarks (Arena Hard, DROP, HumanEval, HumanEval+, MMLU).

Phi 4 significantly outperforms across most benchmarks.

Fri Sep 18 2026 • llm-stats.com

Human preference

Blind head-to-head votes and playground preference scores

Model Size

Parameter count comparison

7.7B diff

Phi 4 has 7.7B more parameters than IBM Granite 4.0 Tiny Preview, making it 110.0% larger.

IBM
IBM Granite 4.0 Tiny Preview
7.0Bparameters
Microsoft
Phi 4
14.7Bparameters
7.0B
IBM Granite 4.0 Tiny Preview
14.7B
Phi 4

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).

IBM
IBM Granite 4.0 Tiny Preview
Input- tokens
Output- tokens
Microsoft
Phi 4
Input16,384 tokens
Output16,384 tokens
Fri Sep 18 2026 • llm-stats.com

License

Usage and distribution terms

IBM Granite 4.0 Tiny Preview is licensed under Apache 2.0, while Phi 4 uses MIT.

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

IBM Granite 4.0 Tiny Preview

Apache 2.0

Open weights

Phi 4

MIT

Open weights

Release Timeline

When each model was launched

IBM Granite 4.0 Tiny Preview was released on 2025-05-02, while Phi 4 was released on 2024-12-12.

IBM Granite 4.0 Tiny Preview is 5 months newer than Phi 4.

IBM Granite 4.0 Tiny Preview

May 2, 2025

1.4 years ago

4mo 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 IBM Granite 4.0 Tiny Preview'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 IBM Granite 4.0 Tiny Preview's cutoff date.

IBM Granite 4.0 Tiny Preview

Phi 4

Jun 2024

Outputs Comparison

Notice missing or incorrect data?Start an Issue discussion

Judge for yourself.

Run your own prompts against IBM Granite 4.0 Tiny Preview and Phi 4 side-by-side, then vote on the output you prefer.

IBM Granite 4.0 Tiny Preview
✓ Preferred
Phi 4
Open in Playground

FAQ

Common questions about IBM Granite 4.0 Tiny Preview vs Phi 4.

Which is better, IBM Granite 4.0 Tiny Preview or Phi 4?

Phi 4 leads the LLM Stats Score 5.4 to -5.5. IBM Granite 4.0 Tiny Preview is made by IBM 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 IBM Granite 4.0 Tiny Preview compare to Phi 4 in benchmarks?

IBM Granite 4.0 Tiny Preview scores AttaQ: 86.1%, HumanEval: 82.4%, HumanEval+: 78.3%, GSM8k: 70.1%, IFEval: 63.0%. 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 IBM Granite 4.0 Tiny Preview and Phi 4?

IBM Granite 4.0 Tiny Preview 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 IBM Granite 4.0 Tiny Preview and Phi 4?

Key differences include LLM Stats Score (-5.5 vs 5.4), licensing (Apache 2.0 vs MIT). See the full comparison above for benchmark-by-benchmark results.

Who makes IBM Granite 4.0 Tiny Preview and Phi 4?

IBM Granite 4.0 Tiny Preview is developed by IBM and Phi 4 is developed by Microsoft.