Currently listed through these providers:
Model details
Qwen3.6 35B-A3B
Qwen3.6 35B-A3B uses a sparse Mixture-of-Experts design with 35 billion total parameters and 3 billion active per token, a configuration intended to improve efficiency without adopting the compute profile of a similarly sized dense model. The architecture supports both thinking and non-thinking operation, giving developers flexibility in how much reasoning to apply to a task.
The model is aimed primarily at agentic coding, where it plans and acts across software-development workflows. Its official positioning says it substantially improves on the preceding Qwen3.5 35B-A3B and competes with larger dense models, while independent reporting records a 73.4% SWE-bench Verified result. That combination makes it a practical fit for teams seeking capable local coding assistance with lower active parameter use, though benchmark claims should be interpreted within the reported test setup.
Quick Info
Powered by- Provider
- SiliconFlow
- Model key
- Qwen/Qwen3.6-35B-A3B
- Release date
- Apr 17, 2026
- Last updated
- Apr 17, 2026
- Input modalities
- Output modalities
- Capabilities
Cost
- Input token cost
- $0.20
- Output token cost
- $1.60
Limits
- Output tokens
- 262,144 tokens
- Context window
- 262,144 tokens