HF Model & Benchmark Matrix

24,078 variants · 14,984 scores · 120 benchmarks
Home/Orgs/Qwen/Qwen/Qwen3-Next-80B-A3B-Instruct

Qwen/Qwen3-Next-80B-A3B-Instruct

family qwen · 13 variants · unknown

provider
Qwen
released
—
huggingface
https://huggingface.co/Qwen/Qwen3-Next-80B-A3B-Instruct
best scores
MMLU-Pro: 78.4MMLU: 89.3GPQA: 70.4AIME 2025: 61.3LiveCodeBench: 43.2IFEval: 84.7Chatbot Arena: 69.0BFCL (function calling): 65.1

Best scores per benchmark

benchmarkcategorybest score
MMLUknowledge89.3
IFEvalinstruction-following84.7
MMLU-Proknowledge78.4
GPQAreasoning70.4
Chatbot Arenachat69.0
BFCL (function calling)agentic65.1
AIME 2025math61.3
LiveCodeBenchcoding43.2

Variants (13)

variantkindquantbitsvRAMdlorgyear
Qwen3-Next-80B-A3B-Instruct-FP8fp8fp88.0302,871Qwen2025
Qwen3-Next-80B-A3B-Instructoriginalbf1616.0288,243Qwen2025
Qwen3-Next-80B-A3B-Instruct-NVFP4nvfp4nvfp44.039,585nvidia2026
Qwen3-Next-80B-A3B-Instruct-GGUFggufgguf4.021,849Qwen2025
Qwen3-Next-80B-A3B-Instruct-GGUFggufgguf16.020,384unsloth2026
Qwen3-Next-80B-A3B-Instruct-MLX-4bitmlxmlx17,726lmstudio-community2025
Qwen3-Next-80B-A3B-Instruct-MLX-8bitmlxmlx16,258lmstudio-community2025
Qwen3-Next-80B-A3B-Instruct-MLX-6bitmlxmlx16,023lmstudio-community2025
Qwen3-Next-80B-A3B-Instruct-MLX-5bitmlxmlx15,938lmstudio-community2025
Qwen3-Next-80B-A3B-Instruct-GGUFggufgguf4.08,155lmstudio-community2026
Qwen_Qwen3-Next-80B-A3B-Instruct-GGUFggufgguf1.03,577bartowski2026
Qwen3-Next-80B-A3B-Instruct-bnb-4bitbnb-4bit4bit908unsloth2025
Qwen3-Next-80B-A3B-Instructoriginalbf1616.0294unsloth2026
Hugging Face model & benchmark data scraped & matrix-ified. 24,078 variants · 14,984 scores · 120 benchmarks.