Compare AI models

Specifications
Parameters
8B0.5B/1.5B/3B/7B/14B/32B
Context window
8k128k
API pricing
Cheapest input
—$0.66Cloudflare
Cheapest output
—$1.00Cloudflare
—$182Cloudflare
Benchmarks
HumanEval
36.22%92.7%
MMLU
69.51%77.6%
GSM8K
58.45%93%
Overview
CompanyNVIDIAQwen
Release dateAug 21 2024Sep 19 2024
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

Qwen2.5-Coder leads Mistral NeMo Minitron 8B on 3 of the 3 benchmarks they both report (HumanEval, MMLU, GSM8K). Mistral NeMo Minitron 8B shipped 29 days before Qwen2.5-Coder, so benchmark comparisons should account for the intervening progress.

Mistral NeMo Minitron 8B has 8B parameters, while Qwen2.5-Coder has 0.5B/1.5B/3B/7B/14B/32B. Context windows are 8k (Mistral NeMo Minitron 8B) vs 128k (Qwen2.5-Coder).

On HumanEval, Qwen2.5-Coder leads at 92.7% vs Mistral NeMo Minitron 8B at 36.22%. On MMLU, Qwen2.5-Coder leads at 77.6% vs Mistral NeMo Minitron 8B at 69.51%. On GSM8K, Qwen2.5-Coder leads at 93% vs Mistral NeMo Minitron 8B at 58.45%.