qwen3-next:80b-cloud

Family context: Qwen3-Next family reference

Publisher: Qwen Team, Alibaba Cloud

Qwen3-Next introduces hybrid Gated DeltaNet/Gated Attention and sparse mixture-of-experts text models. The Instruct card is a specific reference within the family.

Publisher-described context, separate from Sigma5C observations and recorded catalog statistics.

Published family information

Named reference size: 80B total and 3B active parameters, publisher notation Source
Named reference native context: 262,144 tokens Source

Publisher-described uses

Text generation and tool-oriented applications described by Qwen. Source

Reference limitations

The generic site tag does not choose between Instruct and Thinking variants; the Instruct reference must not replace a recorded thinking-capable identity. Context extension is a separate configuration. Source

A recorded tag and an upstream card association do not verify the deployed checkpoint, quantization, configured context or current hosted availability.