Moonlight-16B-A3B-Instruct

Name: Moonlight-16B-A3B-Instruct
Author: Moonshot AI

Chat modelMoonlightMoonlight 16B

Release date: 2025-02-23Updated: 2025-02-23 21:20:35861

Live demoGitHub Hugging Face Compare

Parameters

16B

Context length

Chinese support

Supported

Reasoning ability

Moonlight-16B-A3B-Instruct is an AI model published by Moonshot AI, released on 2025-02-23, for Chat model, with 16B parameters, and 8K context length, requiring about 32GB storage, with a 77.40 score on GSM8K.

Data sourced primarily from official releases (GitHub, Hugging Face, papers), then benchmark leaderboards, then third-party evaluators. Learn about our data methodology

Moonlight-16B-A3B-Instruct

Model basics

Reasoning traces

Not supported

Thinking modes

Thinking modes not supported

Context length

8K tokens

Max output length

No data

Model type

Chat model

Modality (in / out)

Text → Text

Release date

2025-02-23

Model file size

32GB

MoE architecture

Total params / Active params

16B / N/A

Knowledge cutoff

No data

Moonlight-16B-A3B-Instruct

Open source & experience

Code license

MIT License

Weights license

MIT License- Commercial use permitted

GitHub repo

https://github.com/MoonshotAI/Moonlight

Hugging Face

https://huggingface.co/moonshotai/Moonlight-16B-A3B-Instruct

Live demo

No live demo

Moonlight-16B-A3B-Instruct

Official resources

Paper

DataLearnerAI blog

Moonlight-16B-A3B-Instruct

API details

API speed

No data

No public API pricing yet.

Moonlight-16B-A3B-Instruct

Benchmark Results

Moonlight-16B-A3B-Instruct currently shows benchmark results led by GSM8K (18 / 26, score 77.40), BBH (15 / 21, score 65.20), MBPP (21 / 28, score 63.80). This page also consolidates core specs, context limits, and API pricing so you can evaluate the model from benchmark results and deployment constraints together.