Model registry

Frontier models,
on your infrastructure.

Quantized open models, built on cyankiwi AWQ and synced daily from Hugging Face.

Deployment guide

gemma-4-26B-A4B-it Base model Image-Text-to-Text

Google

Optimized by cyankiwi

INT4 16 GB 21M

Qwen3.6-27B Base model Image-Text-to-Text

Alibaba

Optimized by cyankiwi

INT4 20 GB 9.4M INT4 28 GB 1.5M NVFP4 29 GB 10k INT8 37 GB 33k

Qwen3.6-35B-A3B Base model Image-Text-to-Text

Alibaba

Optimized by cyankiwi

INT4 22 GB 6.2M NVFP4 26 GB 22k

gemma-4-31B-it Base model Image-Text-to-Text

Google

Optimized by cyankiwi

INT4 21 GB 4.6M

Qwen3.8-27B Base model Image-Text-to-Text

Alibaba

Optimized by cyankiwi

INT4 19 GB 2.4M INT4 27 GB 44k NVFP4 31 GB 101 FP8 31 GB 11k FP8 23 GB 2

gemma-4-12B-it Base model Any-to-Any

Google

Optimized by cyankiwi

INT4 11 GB 1.5M

gemma-4-E4B-it Base model Any-to-Any

Google

Optimized by cyankiwi

INT4 10 GB 958k INT8 12 GB 24k

GLM-5.2 Base model Text-to-Text

Zhipu

Optimized by cyankiwi

INT4 417 GB 520k

gemma-4-31B-it-qat-q4_0-unquantized Base model Image-Text-to-Text

Google

Optimized by cyankiwi

INT4 21 GB 509k

gemma-4-26B-A4B-it-qat-q4_0-unquantized Base model Image-Text-to-Text

Google

Optimized by cyankiwi

INT4 17 GB 476k

diffusiongemma-26B-A4B-it Base model Image-Text-to-Text

Google

Optimized by cyankiwi

INT4 17 GB 387k

gemma-4-12B-it-qat-q4_0-unquantized Base model Any-to-Any

Google

Optimized by cyankiwi

INT4 11 GB 345k

gemma-4-E2B-it Base model Any-to-Any

Google

Optimized by cyankiwi

INT4 7.6 GB 293k INT8 8.4 GB 4.3k

KAT-Coder-V2.5-Dev Base model Text-to-Text

Kwaipilot

Optimized by cyankiwi

INT4 24 GB 147k

Ornith-1.0-9B Base model Text-to-Text

DeepReinforce

Optimized by cyankiwi

INT4 9 GB 113k FP8 12 GB 30k

Instella-MoE-16B-A3B-Think Base model Text-to-Text

amd

Optimized by cyankiwi

INT4 10 GB 118k

Nex-N2-mini Base model Text-to-Text

nex-agi

Optimized by cyankiwi

INT4 24 GB 112k

Ornith-1.0-35B Base model Text-to-Text

DeepReinforce

Optimized by cyankiwi

INT4 24 GB 40k NVFP4 24 GB 1k FP8 38 GB 56k

ThinkingCap-Qwen3.6-27B Base model Image-Text-to-Text

bottlecapai

Optimized by cyankiwi

INT4 21 GB 67k

MiniMax-M3 Base model Image-Text-to-Text

MiniMax

Optimized by cyankiwi

INT4 258 GB 64k

Llama-3.1-8B-Instruct Base model Text-to-Text

Meta

Optimized by cyankiwi

INT4 6.1 GB 40k

Ornith-1.5-35B-A3B Base model Text-to-Text

ornith-ai

Optimized by cyankiwi

INT4 24 GB 29k FP8 40 GB 11k

granite-4.1-8b Base model Text-to-Text

IBM

Optimized by cyankiwi

INT4 5.4 GB 32k

Mistral-Medium-3.5-128B Base model Text-to-Text

Mistral

Optimized by cyankiwi

INT4 82 GB 28k

Agents-A1 Base model Text-to-Text

InternScience

Optimized by cyankiwi

INT4 24 GB 8.5k NVFP4 24 GB 817 FP8 38 GB 9.2k

Step-3.7-Flash Base model Image-Text-to-Text

StepFun

Optimized by cyankiwi

INT4 138 GB 18k

Qwen-AgentWorld-35B-A3B Base model Text-to-Text

Alibaba

Optimized by cyankiwi

INT4 24 GB 15k NVFP4 23 GB 794 FP8 37 GB 2.1k

Laguna-S-2.1 Base model Text-to-Text

poolside

Optimized by cyankiwi

INT4 68 GB 16k FP8 124 GB 174

gemma-4-E4B-it-qat-q4_0-unquantized Base model Any-to-Any

Google

Optimized by cyankiwi

INT4 10 GB 16k

Llama-3.2-3B-Instruct Base model Text-to-Text

Meta

Optimized by cyankiwi

INT4 2.4 GB 16k

GRM-2.6-Plus Base model Image-Text-to-Text

OrionLLM

Optimized by cyankiwi

INT4 28 GB 12k INT4 20 GB 2.4k

granite-4.1-3b Base model Text-to-Text

IBM

Optimized by cyankiwi

INT4 2.3 GB 14k

Qwen3.8-Flash-Next Base model Image-Text-to-Text

Alibaba

Optimized by cyankiwi

INT4 179 GB 14k

Ornith-1.5-9B Base model Text-to-Text

ornith-ai

Optimized by cyankiwi

INT4 8.5 GB 13k FP8 12 GB 884

Laguna-XS.2 Base model Text-to-Text

poolside

Optimized by cyankiwi

INT4 19 GB 11k

Muse-Glimmer-30B Base model Image-Text-to-Text

meta-models

Optimized by cyankiwi

INT4 22 GB 10k FP8 35 GB 699

GLM-5.3-Flash Base model Image-Text-to-Text

Zhipu

Optimized by cyankiwi

INT4 41 GB 8.8k

LFM2.5-8B-A1B Base model Text-to-Text

Liquid AI

Optimized by cyankiwi

INT4 5.2 GB 7.2k FP8 9.1 GB 1.4k

Hy3 Base model Text-to-Text

Tencent

Optimized by cyankiwi

INT4 182 GB 4.2k NVFP4 177 GB 4.2k

Swift-Qwen3.8-27b Base model Image-Text-to-Text

ukisai

Optimized by cyankiwi

INT4 19 GB 5.8k INT4 27 GB 460

Mellum2-12B-A2.5B-Thinking Base model Text-to-Text

JetBrains

Optimized by cyankiwi

INT4 8.9 GB 4.2k

Swift-1.5-Qwen3.8-27b Base model Image-Text-to-Text

ukisai

Optimized by cyankiwi

INT4 19 GB 2.6k INT4 27 GB 1.2k

MiniCPM5-1B Base model Text-to-Text

OpenBMB

Optimized by cyankiwi

INT4 1.2 GB 3.1k FP8 1.5 GB 507

granite-4.1-30b Base model Text-to-Text

IBM

Optimized by cyankiwi

INT4 20 GB 3.6k

Laguna-XS-2.1 Base model Text-to-Text

poolside

Optimized by cyankiwi

INT4 23 GB 3.1k

North-Mini-Code-1.0 Base model Text-to-Text

Cohere

Optimized by cyankiwi

INT4 18 GB 2.8k

Inkling-Small Base model Image-Text-to-Text

thinkingmachines

Optimized by cyankiwi

INT4 143 GB 2.8k

Llama-3.3-70B-Instruct Base model Text-to-Text

Meta

Optimized by cyankiwi

INT4 44 GB 2.4k

GLM-5.3 Base model Text-to-Text

Zhipu

Optimized by cyankiwi

INT4 432 GB 1.7k

Mellum2-12B-A2.5B-Instruct Base model Text-to-Text

JetBrains

Optimized by cyankiwi

INT4 8.9 GB 1.5k

MiMo-V2.6-Distill-Qwen-9B Base model Image-Text-to-Text

Xiaomi

Optimized by cyankiwi

INT4 8.5 GB 782 FP8 14 GB 71

K2-Horizon-7B Base model Text-to-Text

IFM

Optimized by cyankiwi

INT4 7.6 GB 503 FP8 11 GB 125

Intern-S2-Preview Base model Image-Text-to-Text

InternLM

Optimized by cyankiwi

INT4 23 GB 613

K2-Horizon-MoVA-36B-A4B Base model Text-to-Text

IFM

Optimized by cyankiwi

INT4 21 GB 586

AREX-2 Base model Image-Text-to-Text New

BAAI

Optimized by cyankiwi

INT4 18 GB 505

Spark-X2.5-4B Base model Text-to-Text

XHToken

Optimized by cyankiwi

INT4 2.6 GB 417 FP8 4.4 GB 64

MiniCPM5-2B Base model Text-to-Text

OpenBMB

Optimized by cyankiwi

INT4 2.1 GB 267 FP8 3.1 GB 179

K2-Horizon-3.7B Base model Text-to-Text

IFM

Optimized by cyankiwi

INT4 4.5 GB 241 FP8 6.3 GB 68

ThinkingCap-Qwen3.8-27B Base model Image-Text-to-Text New

bottlecapai

Optimized by cyankiwi

INT4 19 GB 135

Thomson-1.0-Small Base model Image-Text-to-Text

thomsonreuters

Optimized by cyankiwi

INT4 22 GB 66

Apodex-1.1-mini Base model Text-to-Text

apodex

Optimized by cyankiwi

INT4 24 GB 60

Agnes-3.0-Flash Base model Image-Text-to-Text

Agnes-AI

Optimized by cyankiwi

INT4 22 GB 49

Llama-4-Scout-17B-16E-Instruct Base model Image-Text-to-Text

Meta

Optimized by cyankiwi

INT4 76 GB 35

granite-4.2-30b Base model Text-to-Text

IBM

Optimized by cyankiwi

INT4 16 GB 30

LFM2.5-24B-A2B Base model Text-to-Text New

Liquid AI

Optimized by cyankiwi

INT4 13 GB 15

AesCode-32B Base model Image-Text-to-Text New

Microsoft

Optimized by cyankiwi

INT4 20 GB 11

AesCode-8B Base model Image-Text-to-Text New

Microsoft

Optimized by cyankiwi

INT4 7.1 GB 10

Mellum2.1-12B-A2.5B-Thinking Base model Text-to-Text New

JetBrains

Optimized by cyankiwi

INT4 6.8 GB 10

Nex-N2.5-mini Base model Text-to-Text

nex-agi

Optimized by cyankiwi

INT4 26 GB 0

Synced daily from Hugging Face · last sync 2026-10-16

Frontier models,on your infrastructure.

Frontier models,
on your infrastructure.