All Models

Qwen 3.5 35B A3B

qwenReasoningTool CallingAttachmentsOpen WeightsStructured Output

The Qwen3.5 Series 35B-A3B is a native vision-language model designed with a hybrid architecture that integrates linear attention mechanisms and a sparse mixture-of-experts model, achieving higher inference efficiency. Its overall...

Providers20
ReleasedFeb 1, 2026
Input Modalitiestext, image, video, audio
Output Modalitiestext
Tarsk Usecoding

Benchmarks

19.3
Intelligence Index

Available Providers (20)

ProviderModel IDInput CostOutput CostContextMax OutputDocs
OrcaRouterqwen/qwen3.5-35b-a3b$0.06/MTok$0.46/MTok262.1K65.5K
EmpirioLabs AIqwen3-5-35b-a3b$0.06/MTok$0.46/MTok256K64K
Merge Gatewayqwen/qwen3.5-35b-a3b$0.06/MTok$0.46/MTok131.1K32.8K
302.AIqwen3.5-35b-a3b$0.06/MTok$0.46/MTok262.1K65.5K
Requestyqwen3.5-35b-a3b$0.14/MTok$1/MTok262.1K262.1K
Deep InfraQwen/Qwen3.5-35B-A3B$0.14/MTok$1/MTok262.1K81.9K
Kilo Gatewayqwen/qwen3.5-35b-a3b$0.16/MTok$1.30/MTok256K16.4K
NanoGPTqwen/qwen3.5-35b-a3b$0.23/MTok$1.80/MTok260.1K65.5K
SiliconFlow (China)Qwen/Qwen3.5-35B-A3B$0.23/MTok$1.86/MTok262.1K65.5K
SiliconFlowQwen/Qwen3.5-35B-A3B$0.24/MTok$1.80/MTok262.1K262.1K
NovitaAIqwen/qwen3.5-35b-a3b$0.25/MTok$2/MTok262.1K65.5K
Hugging FaceQwen/Qwen3.5-35B-A3B$0.25/MTok$2/MTok262.1K65.5K
Alibabaqwen3.5-35b-a3b$0.25/MTok$2/MTok262.1K65.5K
CoreWeaveQwen/Qwen3.5-35B-A3B$0.25/MTok$1.25/MTok262.1K262.1K
Jalapeno CloudQwen3.5-35B-A3B$0.25/MTok$2/MTok262.1K65.5K
Mixlayerqwen/qwen3.5-35b-a3b$0.25/MTok$1.30/MTok262.1K262.1K
Ofoxqwen/qwen3.5-35b-a3b$0.29/MTok$1.83/MTok256K64K
Ofoxbailian/qwen3.5-35b-a3b$0.29/MTok$1.83/MTok262.1K65.5K
Venice AIqwen3-5-35b-a3b$0.31/MTok$1.25/MTok256K16.4K
OpenRouterqwen/qwen3.5-35b-a3b$0.31/MTok$1.25/MTok262.1K16.4K

Capabilities

Reasoning
Tool Calling
Attachments
Open Weights
Structured Output