Qwen3 Next 80B A3B Instruct

Qwen · released 2025-09-11 · apache-2.0 license

Qwen3-Next-80B-A3B-Instruct is Alibaba's efficiency-focused open-weight MoE (80B total, 3B active, Sept 2025, Apache-2.0) performing on par with the larger 235B-Instruct on several benchmarks: MMLU-Pro 80.6, GPQA 72.9, AIME25 69.5.

Key specs

TypeLocal open-weight
Parameters81.32B total · MoE, — active
Architectureqwen3_next
Context window262K tokens
Knowledge cutoff2025-09-30
Modalitiestext
Recommended backends
Minimum viable rig

Benchmark scores

GPQA Diamond72.9%
SWE-bench Verified
AIME69.5%
MMLU-Pro80.6%
BFCL v3 (tool use)
Composite score7.64
Community ratingNo reviews yet

VRAM & disk per quantization

QuantVRAMDiskRAMContext
Q4_K_M48.7 GB47.2 GB262K

Strengths & weaknesses

Strengths: Highly efficient hybrid-attention ultra-sparse MoE (80B total, 3B active); Near-235B-Instruct quality at a fraction of active params; Strong long context (RULER 91.8)

Weaknesses: Falls short of the 235B on knowledge and some agentic tasks; Non-reasoning mode only