Compare AI models

Specifications
Parameters
—16B
Context window
200k8k
Benchmarks
HumanEval
92%48.1%
GPQA Diamond
59.4%35.2%
MMLU
90.4%70%
GSM8K
96.4%77.4%
Overview
CompanyAnthropicMoonshot AI
Release dateJun 20 2024Feb 22 2025
AccessClosedOpen Weight
Model detailsView modelView model

Frequently asked questions

Claude 3.5 Sonnet leads Moonlight on 4 of the 4 benchmarks they both report (HumanEval, GPQA Diamond, MMLU, GSM8K). Claude 3.5 Sonnet shipped 247 days before Moonlight, so benchmark comparisons should account for the intervening progress.

Context windows are 200k (Claude 3.5 Sonnet) vs 8k (Moonlight). Claude 3.5 Sonnet is closed, while Moonlight is open weight.

On HumanEval, Claude 3.5 Sonnet leads at 92% vs Moonlight at 48.1%. On GPQA Diamond, Claude 3.5 Sonnet leads at 59.4% vs Moonlight at 35.2%. On MMLU, Claude 3.5 Sonnet leads at 90.4% vs Moonlight at 70%. On GSM8K, Claude 3.5 Sonnet leads at 96.4% vs Moonlight at 77.4%.