Qwen3.6-27B
The dense Qwen 3.6 open-weights model with a 27-billion-parameter stack, multimodal inputs, and a 262,144-token native context window.
Qwen3.6-27B is the dense member of the first Qwen 3.6 open-weights release. Every token runs through the full 27-billion-parameter decoder stack rather than selecting a sparse expert subset, which makes it the straightforward comparison point for the MoE operating point in the same family.
At a glance
- Family
- qwen
- Source type
- Open weights
- Released
- April 2026
- Authors
- Qwen Team
- Modalities
- Text, Image, Video
- Parameter count
- 27 billion parameters
- Context length
- 262,144 tokens
What It Is
Qwen3.6-27B is a dense open-weights model from the Qwen Team and the first published Qwen 3.6 checkpoint. Public release materials describe it as a 27-billion-parameter causal language model with a vision encoder, released after the Qwen3.5 series as a stability-focused coding and agentic-workflow upgrade.Inputs And Outputs
The model accepts text, image, and video inputs and produces text outputs. The Hugging Face model card includes Video Input examples alongside text and image prompts, and lists a 262,144-token native context length with longer extension claims noted separately in the same source. This page keeps the native context figure because it is the conservative operating point repeated across official model metadata.Architecture
As a dense operating point, Qwen3.6-27B activates the full parameter stack on each token rather than routing through a sparse expert subset. Official model materials describe a deep decoder with repeated attention and feed-forward blocks, rotary position encoding, and RMS normalization. The public Hugging Face model card also names hybrid attention paths inside the stack; this page's diagram stays at the familiar decoder level—attention, normalization, and a dense feed-forward path—so it teaches the dense flow without turning unexpanded block names into extra claims.Output
ProbabilitiesSoftmaxOutput
LinearTransformer architectureAttention sublayer containerDense feed-forward sublayer containerN×RMSNormDecoder
AttentionRMSNormDense
Feed
ForwardEmbeddingRoPEInput
EmbeddingInput
TokensSoftmax to Output
ProbabilitiesOutput
Linear to SoftmaxRMSNorm to Output
LinearDense
Feed
Forward to RMSNormRMSNorm to Dense
Feed
ForwardRMSNorm to RMSNormDecoder
Attention to RMSNormEmbedding to Decoder
AttentionEmbedding to RMSNormInput
Embedding to EmbeddingRoPE to EmbeddingInput
Tokens to Input
Embedding
Important Modules
The registry-backed modules below are the smallest set worth opening first when you want to connect this dense page to the site's module reference. Rotary position encoding and RMS normalization are named directly in the public model materials and anchor the high-level stack shown above.Training
Official sources describe both pre-training and post-training for the released checkpoint. The Qwen 3.6 release blog emphasizes agentic coding workflows and preserving reasoning context across turns, but this page does not restate training-corpus or recipe details beyond what those primary sources publish.Training regimes
No training regimes listed yet.
Linked papers
No linked paper pages listed yet.