Compare AI models

Specifications
Parameters
8B/70B/405B7B/14B
Context window
128k1M
API pricing
Cheapest input
$0.02DeepInfra—
Cheapest output
$0.04DeepInfra—
$6.00DeepInfra · fp8—
Benchmarks
HumanEval
89%88.4%
MATH
73.8%79.5%
MMLU-Pro
73.3%63.3%
GSM8K
96.8%94.8%
Overview
CompanyMetaQwen
Release dateJul 23 2024Jan 27 2025
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

LLaMA 3.1 leads Qwen2.5-1M on 3 of the 4 benchmarks they both report (HumanEval, MATH, MMLU-Pro, GSM8K). LLaMA 3.1 shipped 188 days before Qwen2.5-1M, so benchmark comparisons should account for the intervening progress.

LLaMA 3.1 has 8B/70B/405B parameters, while Qwen2.5-1M has 7B/14B. Context windows are 128k (LLaMA 3.1) vs 1M (Qwen2.5-1M).

On HumanEval, LLaMA 3.1 leads at 89% vs Qwen2.5-1M at 88.4%. On MATH, Qwen2.5-1M leads at 79.5% vs LLaMA 3.1 at 73.8%. On MMLU-Pro, LLaMA 3.1 leads at 73.3% vs Qwen2.5-1M at 63.3%. On GSM8K, LLaMA 3.1 leads at 96.8% vs Qwen2.5-1M at 94.8%.