Qwen
Qwen3.8-27B
Qwen3.8 27B is a multimodal text, image and video model with configurable thinking. Its text backbone mixes Gated DeltaNet and attention layers, requiring a dedicated memory method.
Specifications
- Source-reported parameters
- 27,781,427,952
- Architectures
- Unknown
- Layers
- 64
- KV heads
- 4
- Head dimension
- 256
- config.json context
- 262144
- Native context in model card
- 262144
- Extended context
- 1000000
- Weight dtype
- Unknown
- File formats
- Unknown
- Declared languages
- Unknown
- Estimation method
- Not validated