🤝 Open to Collab

Pipenetwork PRO

pipenetwork

Kore-Wa's profile picture

ChtiSuisse's profile picture

IamRuuuu's profile picture

https://pipe.network

pipenetwork

AI & ML interests

AI & ML

Recent Activity

updated a collection about 10 hours ago

Ternary-Bonsai-2 MLX

updated a collection about 10 hours ago

Ternary-Bonsai-2 MLX

updated a collection about 10 hours ago

Ternary-Bonsai-2 MLX

View all activity

Organizations

None yet

pipenetwork 's collections 25

Ternary-Bonsai-2 MLX

Stock-runtime MLX builds of Ternary-Bonsai-2-27B: Hadamard rotation unfolded so unmodified mlx-vlm loads it. github.com/PipeNetwork/bonsai2-mlx

pipenetwork/Ternary-Bonsai-2-27B-MLX-bf16

Image-Text-to-Text • 27B • Updated about 10 hours ago
pipenetwork/Ternary-Bonsai-2-27B-MLX-8bit

Image-Text-to-Text • 27B • Updated about 10 hours ago
pipenetwork/Ternary-Bonsai-2-27B-MLX-6bit

Image-Text-to-Text • 27B • Updated about 10 hours ago
pipenetwork/Ternary-Bonsai-2-27B-MLX-4bit

Image-Text-to-Text • 27B • Updated about 10 hours ago • 1

GLM-5.3 MLX

Apple Silicon (MLX) builds of GLM-5.3 (744B) with a fixed glm_moe_dsa runtime: github.com/PipeNetwork/glm53-mlx

pipenetwork/GLM-5.3-MLX-mixed-4_8bit

Text Generation • 743B • Updated 21 days ago • 1.44k • 2
pipenetwork/GLM-5.3-MLX-4bit

Text Generation • 743B • Updated 21 days ago • 876 • 1
pipenetwork/GLM-5.3-MLX-mixed-3_6bit

Text Generation • 743B • Updated 21 days ago • 1.06k • 1
pipenetwork/GLM-5.3-MLX-8bit

Text Generation • 743B • Updated 21 days ago • 478

Qwen3.8-Flash-Next MLX

Apple Silicon (MLX) builds of Qwen3.8-Flash-Next with a validated qwen4_exp runtime: github.com/PipeNetwork/qwen38-flash-next-mlx

pipenetwork/Qwen3.8-Flash-Next-MLX-8bit

Text Generation • 177B • Updated 23 days ago • 2.19k
pipenetwork/Qwen3.8-Flash-Next-MLX-6bit

Text Generation • 177B • Updated 23 days ago • 1.57k • 1
pipenetwork/Qwen3.8-Flash-Next-MLX-mixed-4_8bit

Text Generation • 182B • Updated 23 days ago • 5.95k • 8
pipenetwork/Qwen3.8-Flash-Next-MLX-4bit

Text Generation • 177B • Updated 23 days ago • 1.34k • 1

Muse-Glimmer 30B MLX

Apple Silicon (MLX) build of Muse-Glimmer-30B, plus the runtime no released mlx-vlm has: github.com/PipeNetwork/muse-glimmer-mlx

pipenetwork/Muse-Glimmer-30B-MLX-3bit

Image-Text-to-Text • 30B • Updated Aug 10 • 138
pipenetwork/Muse-Glimmer-30B-MLX-8bit

Image-Text-to-Text • 30B • Updated Aug 10 • 101
pipenetwork/Muse-Glimmer-30B-MLX-4bit

Image-Text-to-Text • 30B • Updated Aug 10 • 114

Inkling-Small · MLX

276B-A12B multimodal MoE on Apple Silicon. Same code path as Inkling 975B. REAP-25 is free: github.com/PipeNetwork/inkling-mlx

pipenetwork/Inkling-Small-MLX-4bit

Image-Text-to-Text • 264B • Updated Jul 30 • 255 • 2
pipenetwork/Inkling-Small-MLX-REAP25-4bit

Image-Text-to-Text • 199B • Updated Jul 30 • 189 • 2
pipenetwork/Inkling-Small-MLX-3bit

Image-Text-to-Text • 264B • Updated Jul 30 • 180 • 1

DeepSeek-V4-Flash · MLX

304B MoE on Apple Silicon. deepseek_v4 is in no released runtime — ported from scratch. Take mixed-4_8bit. github.com/PipeNetwork/deepseek-v4-mlx

pipenetwork/DeepSeek-V4-Flash-MLX-8bit

Text Generation • 284B • Updated Jul 31 • 171
pipenetwork/DeepSeek-V4-Flash-MLX-6bit

Text Generation • 284B • Updated Jul 31 • 93
pipenetwork/DeepSeek-V4-Flash-MLX-mixed-4_8bit

Text Generation • 146B • Updated Jul 31 • 198 • 1
pipenetwork/DeepSeek-V4-Flash-MLX-4bit

Text Generation • 284B • Updated Jul 31 • 160

Qwen3.6-35B-A3B — MLX

MLX (Apple Silicon) conversion of Qwen/Qwen3.6-35B-A3B in NVFP4 — the MLX analog of nvidia/Qwen3.6-35B-A3B-NVFP4. Text-only.

pipenetwork/Qwen3.6-35B-A3B-mlx-nvfp4

Text Generation • 35B • Updated Jun 26 • 31

GLM-5.2 REAP (expert-pruned)

REAP expert-pruned GLM-5.2 (smaller/faster). REAP-25 near-lossless (+2.3% PPL); REAP-37 +7.3%; REAP-50 smallest but +37.5%.

pipenetwork/GLM-5.2-REAP25-MLX-4bit

Text Generation • 562B • Updated 21 days ago • 408 • 2
pipenetwork/GLM-5.2-REAP37-MLX-4bit

Text Generation • 472B • Updated 21 days ago • 373 • 1
pipenetwork/GLM-5.2-REAP50-MLX-4bit

Text Generation • 381B • Updated 21 days ago • 821 • 4

VISTA MLX

First MLX (Apple Silicon) quantizations of inclusionAI VISTA-9B and VISTA-4B (qwen3_5).

pipenetwork/VISTA-9B-MLX-8bit

Text Generation • 9B • Updated Jun 16 • 12
pipenetwork/VISTA-9B-MLX-6bit

Text Generation • 9B • Updated Jun 16 • 11
pipenetwork/VISTA-9B-MLX-5bit

Text Generation • 9B • Updated Jun 16 • 10
pipenetwork/VISTA-9B-MLX-4bit

Text Generation • 9B • Updated Jun 16 • 8

Gemma-4-26B-A4B-it MLX

MLX quantizations of google/gemma-4-26B-A4B-it (MoE): 5/6/8-bit (4-bit available from mlx-community). Text-only.

pipenetwork/Gemma-4-26B-A4B-it-MLX-8bit

Text Generation • 25B • Updated Jun 14 • 31
pipenetwork/Gemma-4-26B-A4B-it-MLX-6bit

Text Generation • 25B • Updated Jun 14 • 23
pipenetwork/Gemma-4-26B-A4B-it-MLX-5bit

Text Generation • 25B • Updated Jun 14 • 24

Kimi-K2.7-Code MLX

MLX build of Kimi-K2.7-Code. Base is natively 4-bit (int4 experts + bf16 rest); this keeps experts at 4-bit and lifts non-expert layers to 6-bit.

pipenetwork/Kimi-K2.7-Code-MLX-4bit-hiprec

Text Generation • 1T • Updated Jun 13 • 1.65k • 1

Holo-3.1 MLX (computer-use)

First working MLX builds of H Company's Holo-3.1 vision-language computer-use agents (Qwen3.5-VL). Vision-validated. Apache-2.0.

pipenetwork/Holo-3.1-4B-MLX-4bit

Image-Text-to-Text • 5B • Updated Jun 5 • 70 • 3
pipenetwork/Holo-3.1-4B-MLX-8bit

Image-Text-to-Text • 5B • Updated Jun 5 • 25 • 1
pipenetwork/Holo-3.1-9B-MLX-4bit

Image-Text-to-Text • 9B • Updated Jun 5 • 21
pipenetwork/Holo-3.1-9B-MLX-8bit

Image-Text-to-Text • 9B • Updated Jun 5 • 28 • 1

Nemotron-3 MLX (Apple Silicon)

MLX quants of NVIDIA Nemotron-3 for Apple Silicon: Ultra 550B (4/5/6/8-bit) and dense Nano-4B (4/8-bit), converted with mlx-lm.

pipenetwork/NVIDIA-Nemotron-3-Ultra-550B-A55B-MLX-4bit

Text Generation • 549B • Updated Jun 5 • 118
pipenetwork/NVIDIA-Nemotron-3-Ultra-550B-A55B-MLX-8bit

Text Generation • 549B • Updated Jun 5 • 109
pipenetwork/NVIDIA-Nemotron-3-Nano-4B-MLX-8bit

1B • Updated Jun 5 • 6
pipenetwork/NVIDIA-Nemotron-3-Nano-4B-MLX-4bit

0.6B • Updated Jun 5 • 6

DeepSeek-V4.1-Flash MLX

Apple Silicon (MLX) builds of DeepSeek-V4.1-Flash (754B) with the only runtime that exists for it: github.com/PipeNetwork/deepseek-v41-mlx

pipenetwork/DeepSeek-V4.1-Flash-MLX-mixed-4_8bit

Text Generation • 379B • Updated 9 days ago • 4.2k • 3
pipenetwork/DeepSeek-V4.1-Flash-MLX-mixed-4_8bit-engram6

Text Generation • 428B • Updated 9 days ago • 1.95k • 1
pipenetwork/DeepSeek-V4.1-Flash-REAP25-MLX-mixed-4_8bit

Text Generation • 311B • Updated 8 days ago • 1.06k
pipenetwork/DeepSeek-V4.1-Flash-REAP50-MLX-mixed-4_8bit

Text Generation • 243B • Updated 8 days ago • 1.07k

GLM-5.3-Flash MLX

Apple Silicon (MLX) builds of GLM-5.3-Flash with a validated, fixed glm5_next runtime: github.com/PipeNetwork/glm53-flash-mlx

pipenetwork/GLM-5.3-Flash-MLX-8bit

Image-Text-to-Text • 314B • Updated 21 days ago • 4.33k • 4
pipenetwork/GLM-5.3-Flash-MLX-6bit

Image-Text-to-Text • 314B • Updated 21 days ago • 6.8k • 3
pipenetwork/GLM-5.3-Flash-MLX-mixed-4_8bit

Image-Text-to-Text • 314B • Updated 21 days ago • 4.06k • 6
pipenetwork/GLM-5.3-Flash-MLX-4bit

Image-Text-to-Text • 314B • Updated 21 days ago • 4.37k • 4

Qwen3.8-2.4T-A95B MLX (REAP)

2.4T-param MoE for Apple Silicon. REAP-pruned because no standard quant fits 550GB. Every build measured per-layer against bf16.

pipenetwork/Qwen3.8-2.4T-A95B-MLX-reap50-3bit

Text Generation • 1.2T • Updated Aug 13 • 748
pipenetwork/Qwen3.8-2.4T-A95B-MLX-reap60-3bit

Text Generation • 998B • Updated Aug 13 • 1.09k • 1
pipenetwork/Qwen3.8-2.4T-A95B-MLX-reap75-4bit

Text Generation • 641B • Updated Aug 13 • 1.3k • 2

MiniMax-H3 MLX

Apple Silicon (MLX) builds of MiniMax-H3, the 33B joint video+audio diffusion transformer. Code: github.com/PipeNetwork/minimax-h3-mlx

pipenetwork/MiniMax-H3-MLX-8bit

Image-Text-to-Video • 9B • Updated Aug 10 • 3.94k • 9
pipenetwork/MiniMax-H3-MLX-6bit

Image-Text-to-Video • 8B • Updated Aug 10 • 450 • 2
pipenetwork/MiniMax-H3-MLX-4bit

Image-Text-to-Video • 7B • Updated Aug 10 • 1.72k • 6
pipenetwork/MiniMax-H3-MLX-bf16

Image-Text-to-Video • 33B • Updated Aug 3 • 971 • 8

Inkling · MLX

975B-A41B multimodal MoE on Apple Silicon. Text+image+audio. Ported from scratch: github.com/PipeNetwork/inkling-mlx

pipenetwork/Inkling-MLX-4bit

Image-Text-to-Text • 947B • Updated Jul 17 • 69 • 1
pipenetwork/Inkling-MLX-REAP12-4bit

Image-Text-to-Text • 835B • Updated Jul 17 • 37
pipenetwork/Inkling-MLX-REAP25-4bit

Image-Text-to-Text • 715B • Updated Jul 17 • 30
pipenetwork/Inkling-MLX-REAP50-4bit

Image-Text-to-Text • 483B • Updated Jul 17 • 33

Nemotron TwoTower · MLX

NVIDIA Nemotron TwoTower 30B-A3B diffusion LM in MLX for Apple Silicon: AR tower + two-tower diffusion, 4/6/8-bit + bf16.

pipenetwork/Nemotron-3-Nano-30B-A3B-context-mlx-4bit

Text Generation • 32B • Updated Jul 2 • 35
pipenetwork/Nemotron-3-Nano-30B-A3B-context-mlx-6bit

Text Generation • 32B • Updated Jul 2 • 33 • 1
pipenetwork/Nemotron-3-Nano-30B-A3B-context-mlx-8bit

Text Generation • 32B • Updated Jul 2 • 29
pipenetwork/Nemotron-3-Nano-30B-A3B-context-mlx-bf16

Text Generation • 32B • Updated Jul 2 • 31

Ornith-1.0-397B — MLX

MLX (Apple Silicon) conversions of deepreinforce-ai/Ornith-1.0-397B (Qwen3.5-MoE, text-only). 4/6/8-bit and bf16.

pipenetwork/Ornith-1.0-397B-mlx-4bit

Text Generation • 396B • Updated Jun 26 • 113 • 3
pipenetwork/Ornith-1.0-397B-mlx-6bit

Text Generation • 396B • Updated Jun 26 • 55
pipenetwork/Ornith-1.0-397B-mlx-8bit

Text Generation • 396B • Updated Jun 26 • 77 • 3

GLM-5.2 MLX

First MLX builds of zai-org/GLM-5.2 (glm_moe_dsa MoE): 4/5/6/8-bit + a 512GB-friendly mixed.

pipenetwork/GLM-5.2-MLX-mixed-3_6bit

Text Generation • 743B • Updated 21 days ago • 1.34k • 6
pipenetwork/GLM-5.2-MLX-4bit

Text Generation • 743B • Updated 21 days ago • 521 • 3
pipenetwork/GLM-5.2-MLX-5bit

Text Generation • 743B • Updated 21 days ago • 486
pipenetwork/GLM-5.2-MLX-6bit

Text Generation • 743B • Updated 21 days ago • 475

Rio-3.1-Open-30B MLX

First MLX (Apple Silicon) quantizations of prefeitura-rio/Rio-3.1-Open-30B (Qwen3-MoE): 4/5/6/8-bit.

pipenetwork/Rio-3.1-Open-30B-MLX-8bit

Text Generation • 31B • Updated Jun 15 • 15
pipenetwork/Rio-3.1-Open-30B-MLX-6bit

Text Generation • 31B • Updated Jun 15 • 11
pipenetwork/Rio-3.1-Open-30B-MLX-5bit

Text Generation • 31B • Updated Jun 15 • 14
pipenetwork/Rio-3.1-Open-30B-MLX-4bit

Text Generation • 31B • Updated Jun 15 • 14

Gemma-4-31B-it MLX

MLX (Apple Silicon) quantizations of google/gemma-4-31B-it: 4/5/6/8-bit. Text-only.

pipenetwork/Gemma-4-31B-it-MLX-8bit

Text Generation • 31B • Updated Jun 13 • 38
pipenetwork/Gemma-4-31B-it-MLX-6bit

Text Generation • 31B • Updated Jun 13 • 21
pipenetwork/Gemma-4-31B-it-MLX-5bit

Text Generation • 31B • Updated Jun 13 • 21
pipenetwork/Gemma-4-31B-it-MLX-4bit

Text Generation • 31B • Updated Jun 13 • 23

MiniMax-M3 MLX

MLX (Apple Silicon) text-only conversions of MiniMax-M3 (427B MoE): 3-bit to 8-bit plus a mixed-precision build.

pipenetwork/MiniMax-M3-MLX-8bit

Text Generation • 426B • Updated Jun 13 • 147 • 1
pipenetwork/MiniMax-M3-MLX-6bit

Text Generation • 426B • Updated Jun 13 • 176 • 1
pipenetwork/MiniMax-M3-MLX-4bit

Text Generation • 426B • Updated Jun 13 • 116
pipenetwork/MiniMax-M3-MLX-mixed-3_6bit

Text Generation • 426B • Updated Jun 13 • 192 • 2

Frog (SWE/debugging) MLX

MLX quants of Microsoft's FrogBoss-32B & FrogMini-14B (Qwen3 debugging finetunes, SWE-bench ~45% pass@1) for Apple Silicon.

pipenetwork/FrogMini-14B-2510-MLX-4bit

Text Generation • 15B • Updated Jun 5 • 14
pipenetwork/FrogMini-14B-2510-MLX-8bit

Text Generation • 15B • Updated Jun 5 • 10
pipenetwork/FrogBoss-32B-2510-MLX-4bit

Text Generation • 33B • Updated Jun 5 • 10
pipenetwork/FrogBoss-32B-2510-MLX-8bit

Text Generation • 33B • Updated Jun 5 • 11

Ternary-Bonsai-2 MLX

Stock-runtime MLX builds of Ternary-Bonsai-2-27B: Hadamard rotation unfolded so unmodified mlx-vlm loads it. github.com/PipeNetwork/bonsai2-mlx

pipenetwork/Ternary-Bonsai-2-27B-MLX-bf16

Image-Text-to-Text • 27B • Updated about 10 hours ago
pipenetwork/Ternary-Bonsai-2-27B-MLX-8bit

Image-Text-to-Text • 27B • Updated about 10 hours ago
pipenetwork/Ternary-Bonsai-2-27B-MLX-6bit

Image-Text-to-Text • 27B • Updated about 10 hours ago
pipenetwork/Ternary-Bonsai-2-27B-MLX-4bit

Image-Text-to-Text • 27B • Updated about 10 hours ago • 1

DeepSeek-V4.1-Flash MLX

Apple Silicon (MLX) builds of DeepSeek-V4.1-Flash (754B) with the only runtime that exists for it: github.com/PipeNetwork/deepseek-v41-mlx

pipenetwork/DeepSeek-V4.1-Flash-MLX-mixed-4_8bit

Text Generation • 379B • Updated 9 days ago • 4.2k • 3
pipenetwork/DeepSeek-V4.1-Flash-MLX-mixed-4_8bit-engram6

Text Generation • 428B • Updated 9 days ago • 1.95k • 1
pipenetwork/DeepSeek-V4.1-Flash-REAP25-MLX-mixed-4_8bit

Text Generation • 311B • Updated 8 days ago • 1.06k
pipenetwork/DeepSeek-V4.1-Flash-REAP50-MLX-mixed-4_8bit

Text Generation • 243B • Updated 8 days ago • 1.07k

GLM-5.3 MLX

Apple Silicon (MLX) builds of GLM-5.3 (744B) with a fixed glm_moe_dsa runtime: github.com/PipeNetwork/glm53-mlx

pipenetwork/GLM-5.3-MLX-mixed-4_8bit

Text Generation • 743B • Updated 21 days ago • 1.44k • 2
pipenetwork/GLM-5.3-MLX-4bit

Text Generation • 743B • Updated 21 days ago • 876 • 1
pipenetwork/GLM-5.3-MLX-mixed-3_6bit

Text Generation • 743B • Updated 21 days ago • 1.06k • 1
pipenetwork/GLM-5.3-MLX-8bit

Text Generation • 743B • Updated 21 days ago • 478

GLM-5.3-Flash MLX

Apple Silicon (MLX) builds of GLM-5.3-Flash with a validated, fixed glm5_next runtime: github.com/PipeNetwork/glm53-flash-mlx

pipenetwork/GLM-5.3-Flash-MLX-8bit

Image-Text-to-Text • 314B • Updated 21 days ago • 4.33k • 4
pipenetwork/GLM-5.3-Flash-MLX-6bit

Image-Text-to-Text • 314B • Updated 21 days ago • 6.8k • 3
pipenetwork/GLM-5.3-Flash-MLX-mixed-4_8bit

Image-Text-to-Text • 314B • Updated 21 days ago • 4.06k • 6
pipenetwork/GLM-5.3-Flash-MLX-4bit

Image-Text-to-Text • 314B • Updated 21 days ago • 4.37k • 4

Qwen3.8-Flash-Next MLX

Apple Silicon (MLX) builds of Qwen3.8-Flash-Next with a validated qwen4_exp runtime: github.com/PipeNetwork/qwen38-flash-next-mlx

pipenetwork/Qwen3.8-Flash-Next-MLX-8bit

Text Generation • 177B • Updated 23 days ago • 2.19k
pipenetwork/Qwen3.8-Flash-Next-MLX-6bit

Text Generation • 177B • Updated 23 days ago • 1.57k • 1
pipenetwork/Qwen3.8-Flash-Next-MLX-mixed-4_8bit

Text Generation • 182B • Updated 23 days ago • 5.95k • 8
pipenetwork/Qwen3.8-Flash-Next-MLX-4bit

Text Generation • 177B • Updated 23 days ago • 1.34k • 1

Qwen3.8-2.4T-A95B MLX (REAP)

2.4T-param MoE for Apple Silicon. REAP-pruned because no standard quant fits 550GB. Every build measured per-layer against bf16.

pipenetwork/Qwen3.8-2.4T-A95B-MLX-reap50-3bit

Text Generation • 1.2T • Updated Aug 13 • 748
pipenetwork/Qwen3.8-2.4T-A95B-MLX-reap60-3bit

Text Generation • 998B • Updated Aug 13 • 1.09k • 1
pipenetwork/Qwen3.8-2.4T-A95B-MLX-reap75-4bit

Text Generation • 641B • Updated Aug 13 • 1.3k • 2

Muse-Glimmer 30B MLX

Apple Silicon (MLX) build of Muse-Glimmer-30B, plus the runtime no released mlx-vlm has: github.com/PipeNetwork/muse-glimmer-mlx

pipenetwork/Muse-Glimmer-30B-MLX-3bit

Image-Text-to-Text • 30B • Updated Aug 10 • 138
pipenetwork/Muse-Glimmer-30B-MLX-8bit

Image-Text-to-Text • 30B • Updated Aug 10 • 101
pipenetwork/Muse-Glimmer-30B-MLX-4bit

Image-Text-to-Text • 30B • Updated Aug 10 • 114

MiniMax-H3 MLX

Apple Silicon (MLX) builds of MiniMax-H3, the 33B joint video+audio diffusion transformer. Code: github.com/PipeNetwork/minimax-h3-mlx

pipenetwork/MiniMax-H3-MLX-8bit

Image-Text-to-Video • 9B • Updated Aug 10 • 3.94k • 9
pipenetwork/MiniMax-H3-MLX-6bit

Image-Text-to-Video • 8B • Updated Aug 10 • 450 • 2
pipenetwork/MiniMax-H3-MLX-4bit

Image-Text-to-Video • 7B • Updated Aug 10 • 1.72k • 6
pipenetwork/MiniMax-H3-MLX-bf16

Image-Text-to-Video • 33B • Updated Aug 3 • 971 • 8

Inkling-Small · MLX

276B-A12B multimodal MoE on Apple Silicon. Same code path as Inkling 975B. REAP-25 is free: github.com/PipeNetwork/inkling-mlx

pipenetwork/Inkling-Small-MLX-4bit

Image-Text-to-Text • 264B • Updated Jul 30 • 255 • 2
pipenetwork/Inkling-Small-MLX-REAP25-4bit

Image-Text-to-Text • 199B • Updated Jul 30 • 189 • 2
pipenetwork/Inkling-Small-MLX-3bit

Image-Text-to-Text • 264B • Updated Jul 30 • 180 • 1

Inkling · MLX

975B-A41B multimodal MoE on Apple Silicon. Text+image+audio. Ported from scratch: github.com/PipeNetwork/inkling-mlx

pipenetwork/Inkling-MLX-4bit

Image-Text-to-Text • 947B • Updated Jul 17 • 69 • 1
pipenetwork/Inkling-MLX-REAP12-4bit

Image-Text-to-Text • 835B • Updated Jul 17 • 37
pipenetwork/Inkling-MLX-REAP25-4bit

Image-Text-to-Text • 715B • Updated Jul 17 • 30
pipenetwork/Inkling-MLX-REAP50-4bit

Image-Text-to-Text • 483B • Updated Jul 17 • 33

DeepSeek-V4-Flash · MLX

304B MoE on Apple Silicon. deepseek_v4 is in no released runtime — ported from scratch. Take mixed-4_8bit. github.com/PipeNetwork/deepseek-v4-mlx

pipenetwork/DeepSeek-V4-Flash-MLX-8bit

Text Generation • 284B • Updated Jul 31 • 171
pipenetwork/DeepSeek-V4-Flash-MLX-6bit

Text Generation • 284B • Updated Jul 31 • 93
pipenetwork/DeepSeek-V4-Flash-MLX-mixed-4_8bit

Text Generation • 146B • Updated Jul 31 • 198 • 1
pipenetwork/DeepSeek-V4-Flash-MLX-4bit

Text Generation • 284B • Updated Jul 31 • 160

Nemotron TwoTower · MLX

NVIDIA Nemotron TwoTower 30B-A3B diffusion LM in MLX for Apple Silicon: AR tower + two-tower diffusion, 4/6/8-bit + bf16.

pipenetwork/Nemotron-3-Nano-30B-A3B-context-mlx-4bit

Text Generation • 32B • Updated Jul 2 • 35
pipenetwork/Nemotron-3-Nano-30B-A3B-context-mlx-6bit

Text Generation • 32B • Updated Jul 2 • 33 • 1
pipenetwork/Nemotron-3-Nano-30B-A3B-context-mlx-8bit

Text Generation • 32B • Updated Jul 2 • 29
pipenetwork/Nemotron-3-Nano-30B-A3B-context-mlx-bf16

Text Generation • 32B • Updated Jul 2 • 31

Qwen3.6-35B-A3B — MLX

MLX (Apple Silicon) conversion of Qwen/Qwen3.6-35B-A3B in NVFP4 — the MLX analog of nvidia/Qwen3.6-35B-A3B-NVFP4. Text-only.

pipenetwork/Qwen3.6-35B-A3B-mlx-nvfp4

Text Generation • 35B • Updated Jun 26 • 31

Ornith-1.0-397B — MLX

MLX (Apple Silicon) conversions of deepreinforce-ai/Ornith-1.0-397B (Qwen3.5-MoE, text-only). 4/6/8-bit and bf16.

pipenetwork/Ornith-1.0-397B-mlx-4bit

Text Generation • 396B • Updated Jun 26 • 113 • 3
pipenetwork/Ornith-1.0-397B-mlx-6bit

Text Generation • 396B • Updated Jun 26 • 55
pipenetwork/Ornith-1.0-397B-mlx-8bit

Text Generation • 396B • Updated Jun 26 • 77 • 3

GLM-5.2 REAP (expert-pruned)

REAP expert-pruned GLM-5.2 (smaller/faster). REAP-25 near-lossless (+2.3% PPL); REAP-37 +7.3%; REAP-50 smallest but +37.5%.

pipenetwork/GLM-5.2-REAP25-MLX-4bit

Text Generation • 562B • Updated 21 days ago • 408 • 2
pipenetwork/GLM-5.2-REAP37-MLX-4bit

Text Generation • 472B • Updated 21 days ago • 373 • 1
pipenetwork/GLM-5.2-REAP50-MLX-4bit

Text Generation • 381B • Updated 21 days ago • 821 • 4

GLM-5.2 MLX

First MLX builds of zai-org/GLM-5.2 (glm_moe_dsa MoE): 4/5/6/8-bit + a 512GB-friendly mixed.

pipenetwork/GLM-5.2-MLX-mixed-3_6bit

Text Generation • 743B • Updated 21 days ago • 1.34k • 6
pipenetwork/GLM-5.2-MLX-4bit

Text Generation • 743B • Updated 21 days ago • 521 • 3
pipenetwork/GLM-5.2-MLX-5bit

Text Generation • 743B • Updated 21 days ago • 486
pipenetwork/GLM-5.2-MLX-6bit

Text Generation • 743B • Updated 21 days ago • 475

VISTA MLX

First MLX (Apple Silicon) quantizations of inclusionAI VISTA-9B and VISTA-4B (qwen3_5).

pipenetwork/VISTA-9B-MLX-8bit

Text Generation • 9B • Updated Jun 16 • 12
pipenetwork/VISTA-9B-MLX-6bit

Text Generation • 9B • Updated Jun 16 • 11
pipenetwork/VISTA-9B-MLX-5bit

Text Generation • 9B • Updated Jun 16 • 10
pipenetwork/VISTA-9B-MLX-4bit

Text Generation • 9B • Updated Jun 16 • 8

Rio-3.1-Open-30B MLX

First MLX (Apple Silicon) quantizations of prefeitura-rio/Rio-3.1-Open-30B (Qwen3-MoE): 4/5/6/8-bit.

pipenetwork/Rio-3.1-Open-30B-MLX-8bit

Text Generation • 31B • Updated Jun 15 • 15
pipenetwork/Rio-3.1-Open-30B-MLX-6bit

Text Generation • 31B • Updated Jun 15 • 11
pipenetwork/Rio-3.1-Open-30B-MLX-5bit

Text Generation • 31B • Updated Jun 15 • 14
pipenetwork/Rio-3.1-Open-30B-MLX-4bit

Text Generation • 31B • Updated Jun 15 • 14

Gemma-4-26B-A4B-it MLX

MLX quantizations of google/gemma-4-26B-A4B-it (MoE): 5/6/8-bit (4-bit available from mlx-community). Text-only.

pipenetwork/Gemma-4-26B-A4B-it-MLX-8bit

Text Generation • 25B • Updated Jun 14 • 31
pipenetwork/Gemma-4-26B-A4B-it-MLX-6bit

Text Generation • 25B • Updated Jun 14 • 23
pipenetwork/Gemma-4-26B-A4B-it-MLX-5bit

Text Generation • 25B • Updated Jun 14 • 24

Gemma-4-31B-it MLX

MLX (Apple Silicon) quantizations of google/gemma-4-31B-it: 4/5/6/8-bit. Text-only.

pipenetwork/Gemma-4-31B-it-MLX-8bit

Text Generation • 31B • Updated Jun 13 • 38
pipenetwork/Gemma-4-31B-it-MLX-6bit

Text Generation • 31B • Updated Jun 13 • 21
pipenetwork/Gemma-4-31B-it-MLX-5bit

Text Generation • 31B • Updated Jun 13 • 21
pipenetwork/Gemma-4-31B-it-MLX-4bit

Text Generation • 31B • Updated Jun 13 • 23

Kimi-K2.7-Code MLX

MLX build of Kimi-K2.7-Code. Base is natively 4-bit (int4 experts + bf16 rest); this keeps experts at 4-bit and lifts non-expert layers to 6-bit.

pipenetwork/Kimi-K2.7-Code-MLX-4bit-hiprec

Text Generation • 1T • Updated Jun 13 • 1.65k • 1

MiniMax-M3 MLX

MLX (Apple Silicon) text-only conversions of MiniMax-M3 (427B MoE): 3-bit to 8-bit plus a mixed-precision build.

pipenetwork/MiniMax-M3-MLX-8bit

Text Generation • 426B • Updated Jun 13 • 147 • 1
pipenetwork/MiniMax-M3-MLX-6bit

Text Generation • 426B • Updated Jun 13 • 176 • 1
pipenetwork/MiniMax-M3-MLX-4bit

Text Generation • 426B • Updated Jun 13 • 116
pipenetwork/MiniMax-M3-MLX-mixed-3_6bit

Text Generation • 426B • Updated Jun 13 • 192 • 2

Holo-3.1 MLX (computer-use)

First working MLX builds of H Company's Holo-3.1 vision-language computer-use agents (Qwen3.5-VL). Vision-validated. Apache-2.0.

pipenetwork/Holo-3.1-4B-MLX-4bit

Image-Text-to-Text • 5B • Updated Jun 5 • 70 • 3
pipenetwork/Holo-3.1-4B-MLX-8bit

Image-Text-to-Text • 5B • Updated Jun 5 • 25 • 1
pipenetwork/Holo-3.1-9B-MLX-4bit

Image-Text-to-Text • 9B • Updated Jun 5 • 21
pipenetwork/Holo-3.1-9B-MLX-8bit

Image-Text-to-Text • 9B • Updated Jun 5 • 28 • 1

Frog (SWE/debugging) MLX

MLX quants of Microsoft's FrogBoss-32B & FrogMini-14B (Qwen3 debugging finetunes, SWE-bench ~45% pass@1) for Apple Silicon.

pipenetwork/FrogMini-14B-2510-MLX-4bit

Text Generation • 15B • Updated Jun 5 • 14
pipenetwork/FrogMini-14B-2510-MLX-8bit

Text Generation • 15B • Updated Jun 5 • 10
pipenetwork/FrogBoss-32B-2510-MLX-4bit

Text Generation • 33B • Updated Jun 5 • 10
pipenetwork/FrogBoss-32B-2510-MLX-8bit

Text Generation • 33B • Updated Jun 5 • 11

Nemotron-3 MLX (Apple Silicon)

MLX quants of NVIDIA Nemotron-3 for Apple Silicon: Ultra 550B (4/5/6/8-bit) and dense Nano-4B (4/8-bit), converted with mlx-lm.

pipenetwork/NVIDIA-Nemotron-3-Ultra-550B-A55B-MLX-4bit

Text Generation • 549B • Updated Jun 5 • 118
pipenetwork/NVIDIA-Nemotron-3-Ultra-550B-A55B-MLX-8bit

Text Generation • 549B • Updated Jun 5 • 109
pipenetwork/NVIDIA-Nemotron-3-Nano-4B-MLX-8bit

1B • Updated Jun 5 • 6
pipenetwork/NVIDIA-Nemotron-3-Nano-4B-MLX-4bit

0.6B • Updated Jun 5 • 6