Chat21B-A3B official

ernie-4-5-21b-a3b

by Baidu

Text MoE post-trained model with 21B total and 3B activated parameters per token and a 131072-token context length.

#chat #instruct #moe

arcflare pull unsloth/ERNIE-4.5-21B-A3B-PT-GGUF:Q4_K_M

[ 01 ] // quickstart

┌─ arcflare — zsh // tty1

[ 02 ] // readme

ernie-4-5-21b-a3b

Text MoE post-trained model with 21B total and 3B activated parameters per token and a 131072-token context length.

Get it

arcflare pull unsloth/ERNIE-4.5-21B-A3B-PT-GGUF:Q4_K_M

arcflare run ernie-4.5-21b-a3b-pt

ArcFlare reads the GGUF header and loads it at the largest context your free VRAM can actually hold. Use arcflare fit to see that before downloading anything else, or arcflare use opencode to code with it.

Highlights

Downloads the Q4_K_M file by default: 13.3 GB. Also published: Q8_0 (23.2 GB), Q5_K_M (15.6 GB), Q3_K_M (10.5 GB)
Trained for a 128K context
Licence: Apache-2.0. Read it before commercial use

Sizes

21b-a3b · 13.3GB

Other sizes live in sibling repos: swap the size in the repo name when pulling.

[ 03 ] // relatedmore →

qwen3-30b-a3b-instruct-2507

Alibaba Qwen

Chat

Qwen3-30B-A3B non-thinking update: 30.5B-parameter MoE with 3.3B active, 128 experts and 256K native context.

size: 30B-A3B
vram: ~20 GB
license: Apache-2.0

83% of a 24 GB card

#chat #instruct #moe

granite-4-0-h-tiny

IBM

Chat

IBM 7B-parameter (1B active) hybrid long-context instruct model with improved instruction following and tool calling.

size: 7B-A1B
vram: ~6 GB
license: Apache-2.0

25% of a 24 GB card

#chat #instruct #moe

lfm2-8b-a1b

Liquid AI

Chat

Liquid AI on-device hybrid MoE with 8.3B total and 1.5B active parameters and a 32,768-token context.

size: 8B-A1B
vram: ~6 GB
license: LFM Open License v1.0

25% of a 24 GB card

#chat #instruct #moe