# AI Chat | New Model Outperforms Deepseek 🔥

**URL:** <https://onehack.st/t/ai-chat-new-model-outperforms-deepseek/300130>\
**Category:** Give-Away and Freebies\
**Tags:** tools, technology, ai\
**Created:** [February 1, 2025, 2:20pm UTC](https://onehack.st/t/ai-chat-new-model-outperforms-deepseek/300130 "2025-02-01T14:20:07Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![Aina](https://onehack.st/user_avatar/onehack.st/aina/32/167469_2.png) [@Aina](https://onehack.st/u/Aina)\
**Post date:** [February 1, 2025, 2:20pm UTC](https://onehack.st/t/ai-chat-new-model-outperforms-deepseek/300130/1 "2025-02-01T14:20:07Z")

</div>

**Ask Qwen, Know More.**

**[Try It](https://chat.qwenlm.ai/)**

Open source model Qwen2.5-Max, which outperforms DeepSeek v3 and relies on MoE (Mixture of Experts) technology, which combines several specialized models to achieve high performance

Most notable features and performance:

• Outperformed DeepSeek V3 and Llama-3 in standard tests:

• Arena-Hard: scored 89.4%, the highest performance among competing models

• GSM8K (Mathematical Reasoning): scored 94.5%.

• C-Eval: scored 92.2%

• CMMLU: scored 91.9%

## ENJOY & HAPPY LEARNING! ❤

---

<div class="post-metadata">

**Author:** ![jorgemv](https://onehack.st/user_avatar/onehack.st/jorgemv/32/132674_2.png) [@jorgemv](https://onehack.st/u/jorgemv)\
**Post date:** [February 1, 2025, 11:32pm UTC](https://onehack.st/t/ai-chat-new-model-outperforms-deepseek/300130/2 "2025-02-01T23:32:20Z")

</div>

Another useful, cool share, SaM! Thanks.