Compare AI models

Specifications
Parameters
16B—
Context window
8k128k
API pricing
Input price
—$2.50
Output price
—$10.00
Cached input price
—$1.25
Cheapest input
—$2.50Azure
Cheapest output
—$10.00Azure
—$1,000OpenAI list price
Benchmarks
HumanEval
48.1%91%
MATH
45.3%76.6%
GPQA Diamond
35.2%49.9%
MMLU
70%87.2%
Overview
CompanyMoonshot AIOpenAI
Release dateFeb 22 2025May 13 2024
AccessOpen WeightClosed
Model detailsView modelView model

Frequently asked questions

GPT-4o leads Moonlight on 4 of the 4 benchmarks they both report (HumanEval, MATH, GPQA Diamond, MMLU). Only GPT-4o has a verified first-party price for an available API: $2.50 per million input tokens and $10.00 per million output tokens. No currently available pay-as-you-go API rate is tracked for Moonlight. GPT-4o shipped 285 days before Moonlight, so benchmark comparisons should account for the intervening progress.

Context windows are 8k (Moonlight) vs 128k (GPT-4o). Moonlight is open weight, while GPT-4o is closed.

On HumanEval, GPT-4o leads at 91% vs Moonlight at 48.1%. On MATH, GPT-4o leads at 76.6% vs Moonlight at 45.3%. On GPQA Diamond, GPT-4o leads at 49.9% vs Moonlight at 35.2%. On MMLU, GPT-4o leads at 87.2% vs Moonlight at 70%.