Import Qwen 3 Next 80B A3B Instruct in OCI Generative AI

The Qwen3 Next 80B A3B Instruct model is now compatible for import in OCI Generative AI.

Qwen3 Next 80B A3B Instruct is an 80-billion-parameter text-to-text Mixture-of-Experts (MoE) language model built on the Qwen3 Next architecture. During inference, the model activates 3 billion parameters, helping improve efficiency while preserving model capacity. It uses a hybrid attention design that combines Gated DeltaNet and Gated Attention for efficient long-context processing, supports context lengths of up to 256,000 tokens, and includes enhancements designed to help improve inference efficiency, model stability, and overall performance.

In OCI Generative AI, use the following details for these models:

Hugging Face model IDs for import: Qwen/Qwen3-Next-80B-A3B-Instruct

Recommended dedicated AI cluster unit shapes:

• A100_80G_X4
• H100_X4
• H200_X4

For the complete list of models compatible for import, see Compatible Models for Import. For available hardware and deployment steps, see Managing Imported Models. For information about the service, see the Generative AI documentation.