Alibaba Releases Qwen3.8-Max: 2.4T Parameter MoE Model

Alibaba's Qwen team has debuted its new Qwen3.8-2.4T-A95B model on Hugging Face, swiftly climbing the trending list. The model boasts a total parameter count of 2.4 trillion (2.4T) and utilizes a Mixture of Experts (MoE) architecture, activating approximately 95 billion (95B) parameters per inference.

Confirmed

The model's page is live on Hugging Face, confirming a total parameter count of 2.4T with 95B active parameters.
It has officially hit the Hugging Face trending list.
The model will feature open weights, with Nebius Token Factory announced as a Day 0 launch partner.

Unconfirmed

Based on leaked configuration details viewed by developer @multimodalart, the model natively supports robust vision and Agent tool-calling capabilities. However, these feature specifics currently stem solely from the leaked data.

Why it matters

As the latest flagship in the Qwen series, its massive 2.4T parameter scale makes it one of the industry's most closely watched ultra-large models. Its open-weights approach and native multimodal/Agent capabilities have generated significant buzz across the AI community.

2026-08-12 ~ 2026-08-13 · 12 related posts

Primary sources

Qwen's New 2.4T MoE Model Tops Hugging Face Trending — Qwen · 2026-08-12 23:17 (UTC+08:00)
Qwen3.8-Max Weights Released with Commercial Revenue Caps — cedric_chee · 2026-08-13 00:11 (UTC+08:00)
Alibaba releases Qwen3.8 Max: 2.4T-param MoE, 1M context, open for commercial use — AdinaYakup · 2026-08-13 00:38 (UTC+08:00)

Alibaba Releases Qwen3.8-2.4T-A95B: 2.4T-Parameter MoE Model — CodeCrusader24 · 2026-08-12
Qwen Releases Qwen3.8-2.4T-A95B Model on Hugging Face — de4dee · 2026-08-12
Qwen3.8-2.4T Model Surfaces on Hugging Face with Agent Capabilities — multimodalart · 2026-08-12
[source] Qwen's New 2.4T MoE Model Tops Hugging Face Trending — Qwen · 2026-08-12
Alibaba's Qwen3.8 Goes Open Weight: 2.4T Parameters with 95B Active — Arindam_1729 · 2026-08-12
Alibaba's Qwen3.8-2.4T-A95B Drops, Autonomously Optimizes Inference Engine — BanghuaZ · 2026-08-12
Qwen3.8-Max Weights Released: 2.4T Total Parameters — Yuchenj_UW · 2026-08-12
[source] Qwen3.8-Max Weights Released with Commercial Revenue Caps — cedric_chee · 2026-08-13
[source] Alibaba releases Qwen3.8 Max: 2.4T-param MoE, 1M context, open for commercial use — AdinaYakup · 2026-08-13
Qwen Releases Qwen3.8-Max: 2.4T Parameters with 1M Context natively — Scobleizer · 2026-08-13
Qwen3.8 2.4T MoE FP8 Quantization Hits Hugging Face Trending — Qwen · 2026-08-13

1 near-duplicate retellings: AdinaYakup