Bluesky Thread

it’s here! a real Qwen3 model

April 28, 2025 View original thread

it’s here! a real Qwen3 model

huggingface.co/Qwen/Qwen3-0...

Qwen/Qwen3-0.6B-FP8 · Hugging Face

We’re on a journey to advance and democratize artificial intelligence through open source and open science.

36 7

blog: qwenlm.github.io/blog/qwen3/

qwenlm.github.io

Qwen3: Think Deeper, Act Faster

QWEN CHAT GitHub Hugging Face ModelScope Kaggle DEMO DISCORD Introduction Today, we are excited to announce the release of Qwen3, the latest addition to the Qwen family of large language models. Our f...

6

j'approve

Due to advancements in model architecture, increase in training data, and more effective training methods, the overall performance of Qwen3 dense base models matches that of Qwen2.5 base models with more parameters. For instance, Qwen3-1.7B/4B/8B/14B/32B-Base performs as well as Qwen2.5-3B/7B/14B/32B/72B-Base, respectively. Notably, in areas like STEM, coding, and reasoning, Qwen3 dense base models even outperform larger Qwen2.5 models. For Qwen3-MoE base models, they achieve similar performance to Qwen2.5 dense base models while using only 10% of the active parameters. This results in significant savings in both training and inference costs.

5

First explicit MCP-callout in a model release, that I'm aware of

They made their own agent framework

github.com/QwenLM/Qwen-...

Agentic Usages

Qwen3 excels in tool calling capabilities. We recommend using Qwen-Agent to make the best use of agentic ability of Qwen3. Qwen-Agent encapsulates tool-calling templates and tool-calling parsers internally, greatly reducing coding complexity.

To define the available tools, you can use the MCP configuration file, use the integrated tool of Qwen-Agent, or integrate other tools by yourself.

7 1

thanks to all the Alibaba staff who stayed up all night to get this to us by 5pm ET

4

it's on ollama now! ollama.com/library/qwen3

Qwen3 is the latest generation of large language models in Qwen series, offering a comprehensive suite of dense and mixture-of-experts (MoE) models.

5

oh, I did not appreciate this but it include both dense and MoE variants,

- 32B = dense
- 30B-A3B = MoE

6

More like this

Qwen3-4B Instruct & Thinking

gemma3n is now open source, available everywhere — huggingface, transformers,...

Qwen3-Next-80B-A3B Base, Instruct & Thinking