The AI arena is free today

Open Superagent

Ministral 3 (14B Base 2512) vs Qwen2.5 14B Instruct

Ministral 3 (14B Base 2512) and Qwen2.5 14B Instruct are closely matched at 3.6 and 3.8 on the LLM Stats Score.

Mistral AI · Alibaba Cloud / Qwen Team · Updated for 2026

Which is better?

Ministral 3 (14B Base 2512) and Qwen2.5 14B Instruct are closely matched on the overall LLM Stats Score at 3.6 and 3.8.

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 Ministral 3 (14B Base 2512)

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

Choose Qwen2.5 14B Instruct

  • you are already invested in the Alibaba Cloud / Qwen Team ecosystem

At a glance

The differences that matter most.

Core performance indexes
3.6
#308
3.8
#306
3.5
#300
3.3
#301
Cost, coverage & limits
Benchmark wins
1 of 2
1 of 2
Input price
— / M
— / M
Output price
— / M
— / M
Context window

Capability indexes

Additional strengths measured across groups of related public benchmarks

2 shared
Index
Ministral 3 (14B Base 2512)
Qwen2.5 14B Instruct
8.8#256
10.6#250
7.6#141
5.7#154
Conservative TrueSkill rating · higher is betterHow scores work

Individual benchmarks

6 reported for Ministral 3 (14B Base 2512) · 16 for Qwen2.5 14B Instruct

2 shared

Ministral 3 (14B Base 2512) outperforms in 1 benchmarks (MMLU-Redux), while Qwen2.5 14B Instruct is better at 1 benchmark (MMLU).

Both models are evenly matched across the benchmarks.

Fri Sep 18 2026 • llm-stats.com

Human preference

Blind head-to-head votes and playground preference scores

Model Size

Parameter count comparison

700.0M diff

Qwen2.5 14B Instruct has 0.7B more parameters than Ministral 3 (14B Base 2512), making it 5.0% larger.

Mistral AI
Ministral 3 (14B Base 2512)
14.0Bparameters
Alibaba Cloud / Qwen Team
Qwen2.5 14B Instruct
14.7Bparameters
14.0B
Ministral 3 (14B Base 2512)
14.7B
Qwen2.5 14B Instruct

Input capabilities

Documented input modalities across available providers

Ministral 3 (14B Base 2512) supports multimodal inputs, whereas Qwen2.5 14B Instruct does not.

Ministral 3 (14B Base 2512) can handle both text and other forms of data like images, making it suitable for multimodal applications.

Ministral 3 (14B Base 2512)

Text
Images
Audio
Video

Qwen2.5 14B Instruct

Text
Images
Audio
Video

License

Usage and distribution terms

Both models are licensed under Apache 2.0.

Both models share the same licensing terms, providing consistent usage rights.

Ministral 3 (14B Base 2512)

Apache 2.0

Open weights

Qwen2.5 14B Instruct

Apache 2.0

Open weights

Release Timeline

When each model was launched

Ministral 3 (14B Base 2512) was released on 2025-12-04, while Qwen2.5 14B Instruct was released on 2024-09-19.

Ministral 3 (14B Base 2512) is 15 months newer than Qwen2.5 14B Instruct.

Ministral 3 (14B Base 2512)

Dec 4, 2025

9 months ago

1.2yr newer
Qwen2.5 14B Instruct

Sep 19, 2024

2.0 years ago

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 Ministral 3 (14B Base 2512) and Qwen2.5 14B Instruct side-by-side, then vote on the output you prefer.

Ministral 3 (14B Base 2512)
✓ Preferred
Qwen2.5 14B Instruct
Open in Playground

FAQ

Common questions about Ministral 3 (14B Base 2512) vs Qwen2.5 14B Instruct.

Which is better, Ministral 3 (14B Base 2512) or Qwen2.5 14B Instruct?

Ministral 3 (14B Base 2512) and Qwen2.5 14B Instruct are closely matched on the LLM Stats Score at 3.6 and 3.8. Ministral 3 (14B Base 2512) is made by Mistral AI and Qwen2.5 14B Instruct is made by Alibaba Cloud / Qwen Team. The best choice depends on your use case — compare their capability indexes, individual benchmarks, pricing, and limits above.

How does Ministral 3 (14B Base 2512) compare to Qwen2.5 14B Instruct in benchmarks?

Ministral 3 (14B Base 2512) scores MMLU-Redux: 82.0%, MMLU: 79.4%, TriviaQA: 74.9%, Multilingual MMLU: 74.2%, MATH (CoT): 67.6%. Qwen2.5 14B Instruct scores GSM8k: 94.8%, HumanEval: 83.5%, MBPP: 82.0%, MATH: 80.0%, MMLU-Redux: 80.0%.

What are the main differences between Ministral 3 (14B Base 2512) and Qwen2.5 14B Instruct?

Key differences include LLM Stats Score (3.6 vs 3.8), multimodal support (yes vs no). See the full comparison above for benchmark-by-benchmark results.

Who makes Ministral 3 (14B Base 2512) and Qwen2.5 14B Instruct?

Ministral 3 (14B Base 2512) is developed by Mistral AI and Qwen2.5 14B Instruct is developed by Alibaba Cloud / Qwen Team.