Qwen3-Next-80B-A3B-Instruct

Overview

Qwen3-Next-80B-A3B-Instruct is the first in the Qwen3-Next series, featuring groundbreaking architectural innovations. It uses Hybrid Attention combining Gated DeltaNet and Gated Attention for efficient ultra-long context modeling, High-Sparsity MoE with 512 experts (10 activated + 1 shared) achieving extreme low activation ratio, and Multi-Token Prediction for improved performance and faster inference. With 80B total parameters and only 3B activated, it outperforms Qwen3-32B-Base with 10% training cost and 10x throughput for 32K+ contexts. The model performs on par with Qwen3-235B-A22B-Instruct-2507 while excelling at ultra-long-context tasks up to 256K tokens (extensible to 1M with YaRN). Architecture: 48 layers, 15T training tokens, hybrid layout of 12*(3*(Gated DeltaNet->MoE)->(Gated Attention->MoE)).

Qwen3-Next-80B-A3B-Instruct was released on September 10, 2025. API access is available through Novita.

Benchmarks

Qwen3-Next-80B-A3B-Instruct Performance Across Datasets

Scores sourced from the model's scorecard, paper, or official blog posts

llm-stats.com - Sun Dec 14 2025

Pricing

Pricing, performance, and capabilities for Qwen3-Next-80B-A3B-Instruct across different providers:

Provider	Input ($/M)	Output ($/M)	Max Input	Max Output	Latency (s)	Throughput	Quantization	Input	Output
Novitabf16	$0.15	$1.50	65.5K	65.5K	—	—	bf16	Text Image Audio Video	Text Image Audio Video

Example Outputs

API Access