Model Comparison

DeepSeek-V2.5 vs Ministral 3 (8B Instruct 2512)

Both models are evenly matched across the benchmarks.

Want to compare interactively?Try the playground

Performance Benchmarks

Comparative analysis across standard metrics

2 benchmarks

DeepSeek-V2.5 outperforms in 1 benchmarks (Arena Hard), while Ministral 3 (8B Instruct 2512) is better at 1 benchmark (MATH).

Both models are evenly matched across the benchmarks.

Tue Apr 21 2026 • llm-stats.com

Arena Performance

Human preference votes

Pricing Analysis

Price comparison per million tokens

Cost data unavailable.

Lowest available price from all providers
Tue Apr 21 2026 • llm-stats.com
DeepSeek
DeepSeek-V2.5
Input tokens$0.14
Output tokens$0.28
Best providerDeepSeek
Mistral AI
Ministral 3 (8B Instruct 2512)
Input tokens$0.00
Output tokens$0.00
Best providerUnknown Organization
Notice missing or incorrect data?Start an Issue→

Model Size

Parameter count comparison

228.0B diff

DeepSeek-V2.5 has 228.0B more parameters than Ministral 3 (8B Instruct 2512), making it 2850.0% larger.

DeepSeek
DeepSeek-V2.5
236.0Bparameters
Mistral AI
Ministral 3 (8B Instruct 2512)
8.0Bparameters
236.0B
DeepSeek-V2.5
8.0B
Ministral 3 (8B Instruct 2512)

Context Window

Maximum input and output token capacity

Only DeepSeek-V2.5 specifies input context (8,192 tokens). Only DeepSeek-V2.5 specifies output context (8,192 tokens).

DeepSeek
DeepSeek-V2.5
Input8,192 tokens
Output8,192 tokens
Mistral AI
Ministral 3 (8B Instruct 2512)
Input- tokens
Output- tokens
Tue Apr 21 2026 • llm-stats.com

Input Capabilities

Supported data types and modalities

Ministral 3 (8B Instruct 2512) supports multimodal inputs, whereas DeepSeek-V2.5 does not.

Ministral 3 (8B Instruct 2512) can handle both text and other forms of data like images, making it suitable for multimodal applications.

DeepSeek-V2.5

Text
Images
Audio
Video

Ministral 3 (8B Instruct 2512)

Text
Images
Audio
Video

License

Usage and distribution terms

DeepSeek-V2.5 is licensed under deepseek, while Ministral 3 (8B Instruct 2512) uses Apache 2.0.

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

DeepSeek-V2.5

deepseek

Open weights

Ministral 3 (8B Instruct 2512)

Apache 2.0

Open weights

Release Timeline

When each model was launched

DeepSeek-V2.5 was released on 2024-05-08, while Ministral 3 (8B Instruct 2512) was released on 2025-12-04.

Ministral 3 (8B Instruct 2512) is 19 months newer than DeepSeek-V2.5.

DeepSeek-V2.5

May 8, 2024

2.0 years ago

Ministral 3 (8B Instruct 2512)

Dec 4, 2025

4 months ago

1.6yr 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→

Key Takeaways

DeepSeek
Larger context window (8,192 tokens)
Higher Arena Hard score (76.2% vs 50.9%)
Mistral AI
Supports multimodal inputs
Higher MATH score (87.6% vs 74.7%)

Detailed Comparison

AI Model Comparison Table
Feature
DeepSeek
DeepSeek-V2.5
Mistral AI
Ministral 3 (8B Instruct 2512)

FAQ

Common questions about DeepSeek-V2.5 vs Ministral 3 (8B Instruct 2512)

Both models are evenly matched across the benchmarks. DeepSeek-V2.5 is made by DeepSeek and Ministral 3 (8B Instruct 2512) is made by Mistral AI. The best choice depends on your use case — compare their benchmark scores, pricing, and capabilities above.
DeepSeek-V2.5 scores GSM8k: 95.1%, MT-Bench: 90.2%, HumanEval: 89.0%, BBH: 84.3%, AlignBench: 80.4%. Ministral 3 (8B Instruct 2512) scores MATH: 87.6%, Wild Bench: 66.8%, Arena Hard: 50.9%, MM-MT-Bench: 8.1%.
DeepSeek-V2.5 supports 8K tokens and Ministral 3 (8B Instruct 2512) supports an unknown number of tokens. A larger context window lets you process longer documents, conversations, or codebases in a single request.
Key differences include multimodal support (no vs yes), licensing (deepseek vs Apache 2.0). See the full comparison above for benchmark-by-benchmark results.
DeepSeek-V2.5 is developed by DeepSeek and Ministral 3 (8B Instruct 2512) is developed by Mistral AI.