Compare AI models

Specifications
Parameters
0.5B/1.5B/3B/7B/14B/32B—
Context window
128k—
API pricing
Cheapest input
$0.66Cloudflare—
Cheapest output
$1.00Cloudflare—
$182Cloudflare—
Benchmarks
HumanEval
92.7%85.7%
MATH
76.4%73%
MMLU-Pro
62.3%72%
MMLU
77.6%86.2%
Overview
CompanyQwenSpaceXSI
Release dateSep 19 2024Aug 13 2024
AccessOpen WeightClosed
Model detailsView modelView model

Frequently asked questions

Qwen2.5-Coder and Grok‑2 mini are evenly matched across the 4 benchmarks they both report (HumanEval, MATH, MMLU-Pro, MMLU). Grok‑2 mini shipped 37 days before Qwen2.5-Coder, so benchmark comparisons should account for the intervening progress.

Qwen2.5-Coder is open weight, while Grok‑2 mini is closed.

On HumanEval, Qwen2.5-Coder leads at 92.7% vs Grok‑2 mini at 85.7%. On MATH, Qwen2.5-Coder leads at 76.4% vs Grok‑2 mini at 73%. On MMLU-Pro, Grok‑2 mini leads at 72% vs Qwen2.5-Coder at 62.3%. On MMLU, Grok‑2 mini leads at 86.2% vs Qwen2.5-Coder at 77.6%.