Alibaba Cloud·Qwen

Qwen3.8-Flash-Next

A Qwen4 architecture preview: 125B main model plus 51B N-gram embeddings, 6B active per token, extendable to 1M context.

CurrentReasoningCodingAgentMultimodalOpen weights

MODEL NUMBER076

Specifications

Only published specifications are shown.

Total parameters

176B

Active parameters

Architecture

Hybrid

Context

262K

Input modalities

Text · Image

Open weights

Yes

License

Apache-2.0

Status

Current

A Qwen4 architecture preview: 125B main model plus 51B N-gram embeddings, 6B active per token, extendable to 1M context.

It introduced QSA, Gated Residual, and N-gram memory to reduce compute while advancing long-context agents.

PreviousQwen3.8-Max2.4T · 1M ctx

CurrentQwen3.8-Flash-Next176B · 262K ctx

New capabilities Adjustable reasoning

Gated DeltaNetQwen Sparse Attention

Adjustable reasoningCodingAgentLong context

2026-08-26Open weights

Qwen3.8-Flash-Next released open weights and became available as managed Qwen3.8-Flash.

Select a card to view the related model.