Which is better, Llama 3.1 Instruct 405B or Qwen3.8 Max Preview?

Refresh Models —Deals —Regulatory —Sources —

Llama 3.1 Instruct 405B vs Qwen3.8 Max Preview

15 SHARED BENCHMARKS

Across 15 shared benchmarks, Llama 3.1 Instruct 405B scores higher on 0 and Qwen3.8 Max Preview on 15. The widest gap is AA Agentic Index, where Qwen3.8 Max Preview scores 58.4 against 6.3.

METAVSALIBABA15 SHARED0–15 HEAD-TO-HEAD

BenchmarkLlama 3.1 Instruct 405BMARGINQwen3.8 Max Preview

AA Agentic Index6.358.4

AA Intelligence8.558.1

AA-LCR24.774.3

AA-Omniscience-17.13.4

Artificial Analysis Coding Index14.571.8

critpt020

gdpval061

GPQA Diamond51.592.7

HLE4.243.6

IFBench3982.8

longbench_v236.166.3

OmniScience Accuracy23.231.9

OmniScience Non-Hallucination4958.3

scicode29.952.9

SWE-bench Pro11.267.7

Best tracked score per model per benchmark (default configuration; source-attributed). ↓ marks lower-is-better metrics. Open either model for its full surface, provenance and pricing.