# Qwen3.8-2.4T-A95B MoE model published

> Qwen published a new large-scale MoE text-generation model on Hugging Face: Qwen3.8-2.4T-A95B. The model has 2.4 trillion total parameters with 95 billion active parameters, using the Qwen3.5 MoE architecture. It is a conversational text-generation model compatible with the transformers library and endpoints.

| | |
|---|---|
| **Tool** | Qwen |
| **Version** | — |
| **Kind** | release |
| **Published** | 2026-08-08 |
| **Observed** | 2026-08-21 |
| **Significance** | 1/5 |
| **Breaking** | no |
| **Categories** | — |

> **Note:** this classification is below our confidence threshold and is pending human review.

## What changed


- 2.4T total parameters, 95B active parameters (MoE)
- Architecture: qwen3_5_moe_text
- Pipeline: text-generation via transformers
- Format: safetensors
- Tags: conversational, endpoints_compatible, eval-results, region:us
- License: other (license not specified in the standard fields)


## Sources

- [hf_model](https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B) — retrieved 2026-08-21


## Community

_No curated reactions recorded. Facts and community takes are kept in separate layers
and never blended._

---
Canonical: https://changelogs.info/qwen/qwen3-8-2-4t-a95b-moe-model-published
Entity: https://changelogs.info/qwen
Event ID: `evt_2026-08-08_qwen_qwen-qwen3-8-2-4t-a95b`
Licence: event synthesis © changelogs.info, CC BY 4.0. Linked sources belong to their vendors.