The AI arena is free today

Open Superagent

Gemma 2 27B vs Ministral 3 (3B Base 2512)

Gemma 2 27B significantly outperforms across most benchmarks.

Google · Mistral AI · Updated for 2026

Which is better?

Gemma 2 27B outperforms in 3 benchmarks (AGIEval, MMLU, TriviaQA), while Ministral 3 (3B Base 2512) is better at 0 benchmarks. Gemma 2 27B significantly outperforms across most benchmarks.

Based on current benchmark, pricing, and model metadata for 2026.

Choose Gemma 2 27B

  • you want the strongest raw capability — it leads on 3 of 3 shared benchmarks

Choose Ministral 3 (3B Base 2512)

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

At a glance

The differences that matter most.

Gemma 2 27B
Google
Ministral 3 (3B Base 2512)
Mistral AI
Benchmark wins
3 of 3
0 of 3
Input price
— / M
— / M
Output price
— / M
— / M
Context window
—
—
Released
Jun 2024
Dec 2025
License
Gemma
Apache 2.0

Performance Benchmarks

Comparative analysis across standard metrics

3 benchmarks

Gemma 2 27B outperforms in 3 benchmarks (AGIEval, MMLU, TriviaQA), while Ministral 3 (3B Base 2512) is better at 0 benchmarks.

Gemma 2 27B significantly outperforms across most benchmarks.

Fri Aug 28 2026 • llm-stats.com

Arena Performance

Playground indexes and blind preference scores

Model Size

Parameter count comparison

24.2B diff

Gemma 2 27B has 24.2B more parameters than Ministral 3 (3B Base 2512), making it 806.7% larger.

Google
Gemma 2 27B
27.2Bparameters
Mistral AI
Ministral 3 (3B Base 2512)
3.0Bparameters
27.2B
Gemma 2 27B
3.0B
Ministral 3 (3B Base 2512)

Input Capabilities

Supported data types and modalities

Ministral 3 (3B Base 2512) supports multimodal inputs, whereas Gemma 2 27B does not.

Ministral 3 (3B Base 2512) can handle both text and other forms of data like images, making it suitable for multimodal applications.

Gemma 2 27B

Text
Images
Audio
Video

Ministral 3 (3B Base 2512)

Text
Images
Audio
Video

License

Usage and distribution terms

Gemma 2 27B is licensed under Gemma, while Ministral 3 (3B Base 2512) uses Apache 2.0.

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

Gemma 2 27B

Gemma

Open weights

Ministral 3 (3B Base 2512)

Apache 2.0

Open weights

Release Timeline

When each model was launched

Gemma 2 27B was released on 2024-06-27, while Ministral 3 (3B Base 2512) was released on 2025-12-04.

Ministral 3 (3B Base 2512) is 18 months newer than Gemma 2 27B.

Gemma 2 27B

Jun 27, 2024

2.2 years ago

Ministral 3 (3B Base 2512)

Dec 4, 2025

8 months ago

1.4yr newer

Knowledge Cutoff

When training data ends

Neither model specifies a knowledge cutoff date.

Unable to compare the recency of their training data.

No cutoff dates available

Outputs Comparison

Notice missing or incorrect data?Start an Issue discussion→

Judge for yourself.

Run your own prompts against Gemma 2 27B and Ministral 3 (3B Base 2512) side-by-side, then vote on the output you prefer.

Gemma 2 27B
✓ Preferred
Ministral 3 (3B Base 2512)
Open in Playground

FAQ

Common questions about Gemma 2 27B vs Ministral 3 (3B Base 2512).

Which is better, Gemma 2 27B or Ministral 3 (3B Base 2512)?

Gemma 2 27B significantly outperforms across most benchmarks. Gemma 2 27B is made by Google and Ministral 3 (3B Base 2512) is made by Mistral AI. The best choice depends on your use case — compare their benchmark scores, pricing, and capabilities above.

How does Gemma 2 27B compare to Ministral 3 (3B Base 2512) in benchmarks?

Gemma 2 27B scores ARC-E: 88.6%, HellaSwag: 86.4%, BoolQ: 84.8%, TriviaQA: 83.7%, Winogrande: 83.7%. Ministral 3 (3B Base 2512) scores MMLU-Redux: 73.5%, MMLU: 70.7%, Multilingual MMLU: 65.2%, MATH (CoT): 60.1%, TriviaQA: 59.2%.

What are the main differences between Gemma 2 27B and Ministral 3 (3B Base 2512)?

Key differences include multimodal support (no vs yes), licensing (Gemma vs Apache 2.0). See the full comparison above for benchmark-by-benchmark results.

Who makes Gemma 2 27B and Ministral 3 (3B Base 2512)?

Gemma 2 27B is developed by Google and Ministral 3 (3B Base 2512) is developed by Mistral AI.

Related comparisons