GGUF quants of Nanbeige4.3-33B by AtomicChat
AI & ML interests
Local Models Server and Free Ai Chat app
Recent Activity
View all activity
Atomic Dynamic GGUF of Nemotron 3.5 Bolt E4B, Super VL 120B-A12B and Bolt 74B-A9B, each measured against other GGUF builds on one stand.
-
AtomicChat/NVIDIA-Nemotron-3.5-Bolt-E4B-GGUF
Text Generation • 8B • Updated • 2.65k • 1 -
AtomicChat/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-GGUF
Image-Text-to-Text • 124B • Updated • 3.71k • 5 -
AtomicChat/NVIDIA-Nemotron-3.5-Bolt-74B-A9B-GGUF
Image-Text-to-Text • 74B • Updated • 8.32k • 8 -
AtomicChat/NVIDIA-Nemotron-3.5-Bolt-E4B-GGUF-metrics
Viewer • Updated • 50 • 130
RecursiveSI's Flint (39.8B, 2.5B active) quantized to GGUF by AtomicChat, with its metrics.
BFL's 16-step image and edit model in GGUF: measured AD quants, text encoder, prompt upsampler, and the renders behind them.
DeepSeek-V4.1-Flash-1017 without refusals: GGUF builds, the adapter and their measurements
Atomic Dynamic GGUF quants of Reflection Beam 1.0 (501B-A23B MoE), measured against the other GGUF builds on one stand.
Ternary Bonsai 2 without refusals: rank-1 LoRA adapters for PrismML's 2B, 4B, 9B and 27B packs, which stay byte-identical, plus the metrics.
-
AtomicChat/Ternary-Bonsai-2-27B-Uncensored-LoRA-GGUF
Text Generation • 2.42M • Updated • 1.48k • 12 -
AtomicChat/Ternary-Bonsai-2-9B-Uncensored-LoRA-GGUF
Text Generation • 786k • Updated • 731 • 6 -
AtomicChat/Ternary-Bonsai-2-4B-Uncensored-LoRA-GGUF
Text Generation • 590k • Updated • 280 • 3 -
AtomicChat/Ternary-Bonsai-2-2B-Uncensored-LoRA-GGUF
Text Generation • 295k • Updated • 264 • 3
Atomic Dynamic GGUF quants of inclusionAI Ling 3.1 Flash, measured against the other GGUF builds on one stand.
Atomic Chat self-quantized GGUFs of Liquid AI's LFM2.5 models, built from the original weights with an importance matrix.
OpenBMB's MiniCPM-V 4.7 (35B-A3B MoE and 1B) in GGUF: measured AD quants, projector, and the metrics behind them.
-
AtomicChat/MiniCPM-V-4.7-35B-A3B-GGUF
Image-Text-to-Text • 35B • Updated • 20.5k • 17 -
AtomicChat/MiniCPM-V-4.7-35B-A3B-GGUF-metrics
Viewer • Updated • 615 • 293 -
AtomicChat/MiniCPM-V-4.7-1B-GGUF
Image-Text-to-Text • 1B • Updated • 11.2k • 7 -
AtomicChat/MiniCPM-V-4.7-1B-GGUF-metrics
Viewer • Updated • 283 • 183
Qwen's 8-step image model in GGUF for stable-diffusion.cpp: measured AD quants and the renders behind them.
Liquid AI d1 decision models (d1-3B, d1-omni-600M) in GGUF: measured AD quants and the metrics behind them.
NVFP4 experts for Blackwell, calibrated by NVIDIA's recipe and measured against the original: checkpoint plus the metrics.
-
AtomicChat/Qwen3.8-27B-GGUF
Image-Text-to-Text • 27B • Updated • 61.2k • 179 -
AtomicChat/Qwen3.8-27B-GGUF-metrics
Updated • 2.51k • 4 -
AtomicChat/Qwen3.8-27B-MLX-AD-3.50bpw-DWQ-12.7GB
Image-Text-to-Text • 27B • Updated • 218 • 2 -
AtomicChat/Qwen3.8-27B-MLX-AD-3.70bpw-DWQ-13.4GB
Image-Text-to-Text • 27B • Updated • 287 • 3
Atomic Chat GGUF builds of Google Gemma 4
Inkling by Thinking Machines, self-quantized to GGUF by Atomic Chat. 975B MoE (41B active), 1M context.
DFlash block-diffusion draft heads (GGUF). Attach to any target GGUF to speed up llama.cpp.
Atomic Chat self-quantized GGUFs of Qwen3.6, built from the original weights with an importance matrix.
Microsoft Phi-4 models, self-quantized to GGUF by Atomic Chat with a per-tensor importance matrix.
Atomic Chat self-quantized GGUFs of Qwen3.5, built from the original weights with an importance matrix.
Dynamic-imatrix GGUF quants of Qwen 3.6 27B & 35B-A3B. TurboQuant3 KV + shared-model NextN ready.
Nemotron 3.5 Bolt E4B, Super VL and Bolt 74B with the refusal direction projected out, baked into Atomic Dynamic GGUF builds.
-
AtomicChat/NVIDIA-Nemotron-3.5-Bolt-E4B-Uncensored-GGUF
Text Generation • 864k • Updated • 1.96k • 13 -
AtomicChat/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-Uncensored-GGUF
Image-Text-to-Text • 124B • Updated • 1.84k • 5 -
AtomicChat/NVIDIA-Nemotron-3.5-Bolt-74B-A9B-Uncensored-GGUF
Image-Text-to-Text • 74B • Updated • 9.85k • 12 -
AtomicChat/NVIDIA-Nemotron-3.5-Bolt-E4B-Uncensored-GGUF-metrics
Viewer • Updated • 1.93k • 84
MiniMax M3.1 Flash (428B-A23B) as GGUF for llama.cpp, released on day 0.
FLUX 3 Image Klein's text encoder and prompt upsampler without refusals, as GGUF, with what they do to the pictures measured.
Ornith 1.5 9B, 35B-A3B and 122B quantized to GGUF by AtomicChat, each with its metrics dataset.
-
AtomicChat/Ornith-1.5-122B-GGUF
Image-Text-to-Text • 125B • Updated • 8.64k • 13 -
AtomicChat/Ornith-1.5-35B-A3B-GGUF
Image-Text-to-Text • 35B • Updated • 20.1k • 63 -
AtomicChat/Ornith-1.5-35B-A3B-GGUF-metrics
Updated • 550 • 1 -
AtomicChat/Ornith-1.5-9B-GGUF
Image-Text-to-Text • 9B • Updated • 32.7k • 79
Atomic Dynamic GGUF quants of Meta Muse Spark Open, measured against the other GGUF builds on one stand.
Uncensored builds of Ling 3.1 Flash: refusal direction removed, measured for MMLU and drift.
LFM2.5-24B-A2B without refusals: baked GGUF builds and a LoRA adapter for our original quants, plus the metrics.
MiniCPM-V-4.7-35B-A3B without refusals: baked GGUF builds and a LoRA adapter for our original quants, plus the metrics.
Qwen-Image's text encoder (Qwen3-VL-8B) with the refusal direction removed, a drop-in --llm for sd.cpp, and what it does to images.
Qwen3.8-Flash-Next without refusals: baked GGUF builds and a LoRA adapter for our original quants, plus the metrics.
Google's multimodal embedding model in GGUF: measured AD quants, projector, and the metrics behind them.
Qwen3.8-Flash-Next (177B MoE) in GGUF: measured AD quants and the metrics behind them.
Atomic Chat GGUF builds of Tencent Hy3 (295B MoE): coding-optimized imatrix quants
poolside Laguna XS 2.1 (33B/3B MoE) quantized by Atomic Chat - GGUF (llama.cpp) and MLX (Apple Silicon) ladders.
-
AtomicChat/Laguna-XS-2.1-GGUF
Text Generation • 33B • Updated • 1.47k • 17 -
AtomicChat/Laguna-XS-2.1-MLX-8bit
Text Generation • 33B • Updated • 512 • 2 -
AtomicChat/Laguna-XS-2.1-MLX-6bit
Text Generation • 33B • Updated • 161 • 2 -
AtomicChat/Laguna-XS-2.1-MLX-5bit
Text Generation • 33B • Updated • 263 • 3
Qwen2.5 instruct models, self-quantized to GGUF by Atomic Chat with a per-tensor importance matrix.
DeepReinforce's Ornith 1.0 agentic-coding models, self-quantized to GGUF by Atomic Chat with a per-tensor importance matrix. Runs fully offline.
Atomic Chat self-quantized GGUFs of the Qwen3 Coder models, built from the original weights with an importance matrix.
Gemma 4 MTP assistant drafters as GGUF (F16/Q8_0/Q5_K_M/Q4_K_M/Q4_K_S). Speculative-decoding heads for the atomic-llama-cpp-turboquant fork.
-
AtomicChat/gemma-4-E2B-it-assistant-GGUF
Text Generation • 77.2M • Updated • 1.29k • 8 -
AtomicChat/gemma-4-E4B-it-assistant-GGUF
Text Generation • 78M • Updated • 1.52k • 26 -
AtomicChat/gemma-4-26B-A4B-it-assistant-GGUF
Text Generation • 0.4B • Updated • 2.58k • 33 -
AtomicChat/gemma-4-31B-it-assistant-GGUF
Text Generation • 0.5B • Updated • 1.95k • 22
GGUF quants of Nanbeige4.3-33B by AtomicChat
Nemotron 3.5 Bolt E4B, Super VL and Bolt 74B with the refusal direction projected out, baked into Atomic Dynamic GGUF builds.
-
AtomicChat/NVIDIA-Nemotron-3.5-Bolt-E4B-Uncensored-GGUF
Text Generation • 864k • Updated • 1.96k • 13 -
AtomicChat/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-Uncensored-GGUF
Image-Text-to-Text • 124B • Updated • 1.84k • 5 -
AtomicChat/NVIDIA-Nemotron-3.5-Bolt-74B-A9B-Uncensored-GGUF
Image-Text-to-Text • 74B • Updated • 9.85k • 12 -
AtomicChat/NVIDIA-Nemotron-3.5-Bolt-E4B-Uncensored-GGUF-metrics
Viewer • Updated • 1.93k • 84
Atomic Dynamic GGUF of Nemotron 3.5 Bolt E4B, Super VL 120B-A12B and Bolt 74B-A9B, each measured against other GGUF builds on one stand.
-
AtomicChat/NVIDIA-Nemotron-3.5-Bolt-E4B-GGUF
Text Generation • 8B • Updated • 2.65k • 1 -
AtomicChat/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-GGUF
Image-Text-to-Text • 124B • Updated • 3.71k • 5 -
AtomicChat/NVIDIA-Nemotron-3.5-Bolt-74B-A9B-GGUF
Image-Text-to-Text • 74B • Updated • 8.32k • 8 -
AtomicChat/NVIDIA-Nemotron-3.5-Bolt-E4B-GGUF-metrics
Viewer • Updated • 50 • 130
MiniMax M3.1 Flash (428B-A23B) as GGUF for llama.cpp, released on day 0.
RecursiveSI's Flint (39.8B, 2.5B active) quantized to GGUF by AtomicChat, with its metrics.
FLUX 3 Image Klein's text encoder and prompt upsampler without refusals, as GGUF, with what they do to the pictures measured.
BFL's 16-step image and edit model in GGUF: measured AD quants, text encoder, prompt upsampler, and the renders behind them.
DeepSeek-V4.1-Flash-1017 without refusals: GGUF builds, the adapter and their measurements
Ornith 1.5 9B, 35B-A3B and 122B quantized to GGUF by AtomicChat, each with its metrics dataset.
-
AtomicChat/Ornith-1.5-122B-GGUF
Image-Text-to-Text • 125B • Updated • 8.64k • 13 -
AtomicChat/Ornith-1.5-35B-A3B-GGUF
Image-Text-to-Text • 35B • Updated • 20.1k • 63 -
AtomicChat/Ornith-1.5-35B-A3B-GGUF-metrics
Updated • 550 • 1 -
AtomicChat/Ornith-1.5-9B-GGUF
Image-Text-to-Text • 9B • Updated • 32.7k • 79
Atomic Dynamic GGUF quants of Reflection Beam 1.0 (501B-A23B MoE), measured against the other GGUF builds on one stand.
Atomic Dynamic GGUF quants of Meta Muse Spark Open, measured against the other GGUF builds on one stand.
Ternary Bonsai 2 without refusals: rank-1 LoRA adapters for PrismML's 2B, 4B, 9B and 27B packs, which stay byte-identical, plus the metrics.
-
AtomicChat/Ternary-Bonsai-2-27B-Uncensored-LoRA-GGUF
Text Generation • 2.42M • Updated • 1.48k • 12 -
AtomicChat/Ternary-Bonsai-2-9B-Uncensored-LoRA-GGUF
Text Generation • 786k • Updated • 731 • 6 -
AtomicChat/Ternary-Bonsai-2-4B-Uncensored-LoRA-GGUF
Text Generation • 590k • Updated • 280 • 3 -
AtomicChat/Ternary-Bonsai-2-2B-Uncensored-LoRA-GGUF
Text Generation • 295k • Updated • 264 • 3
Uncensored builds of Ling 3.1 Flash: refusal direction removed, measured for MMLU and drift.
Atomic Dynamic GGUF quants of inclusionAI Ling 3.1 Flash, measured against the other GGUF builds on one stand.
LFM2.5-24B-A2B without refusals: baked GGUF builds and a LoRA adapter for our original quants, plus the metrics.
Atomic Chat self-quantized GGUFs of Liquid AI's LFM2.5 models, built from the original weights with an importance matrix.
MiniCPM-V-4.7-35B-A3B without refusals: baked GGUF builds and a LoRA adapter for our original quants, plus the metrics.
OpenBMB's MiniCPM-V 4.7 (35B-A3B MoE and 1B) in GGUF: measured AD quants, projector, and the metrics behind them.
-
AtomicChat/MiniCPM-V-4.7-35B-A3B-GGUF
Image-Text-to-Text • 35B • Updated • 20.5k • 17 -
AtomicChat/MiniCPM-V-4.7-35B-A3B-GGUF-metrics
Viewer • Updated • 615 • 293 -
AtomicChat/MiniCPM-V-4.7-1B-GGUF
Image-Text-to-Text • 1B • Updated • 11.2k • 7 -
AtomicChat/MiniCPM-V-4.7-1B-GGUF-metrics
Viewer • Updated • 283 • 183
Qwen-Image's text encoder (Qwen3-VL-8B) with the refusal direction removed, a drop-in --llm for sd.cpp, and what it does to images.
Qwen's 8-step image model in GGUF for stable-diffusion.cpp: measured AD quants and the renders behind them.
Qwen3.8-Flash-Next without refusals: baked GGUF builds and a LoRA adapter for our original quants, plus the metrics.
Liquid AI d1 decision models (d1-3B, d1-omni-600M) in GGUF: measured AD quants and the metrics behind them.
Google's multimodal embedding model in GGUF: measured AD quants, projector, and the metrics behind them.
NVFP4 experts for Blackwell, calibrated by NVIDIA's recipe and measured against the original: checkpoint plus the metrics.
Qwen3.8-Flash-Next (177B MoE) in GGUF: measured AD quants and the metrics behind them.
-
AtomicChat/Qwen3.8-27B-GGUF
Image-Text-to-Text • 27B • Updated • 61.2k • 179 -
AtomicChat/Qwen3.8-27B-GGUF-metrics
Updated • 2.51k • 4 -
AtomicChat/Qwen3.8-27B-MLX-AD-3.50bpw-DWQ-12.7GB
Image-Text-to-Text • 27B • Updated • 218 • 2 -
AtomicChat/Qwen3.8-27B-MLX-AD-3.70bpw-DWQ-13.4GB
Image-Text-to-Text • 27B • Updated • 287 • 3
Atomic Chat GGUF builds of Google Gemma 4
Inkling by Thinking Machines, self-quantized to GGUF by Atomic Chat. 975B MoE (41B active), 1M context.
Atomic Chat GGUF builds of Tencent Hy3 (295B MoE): coding-optimized imatrix quants
DFlash block-diffusion draft heads (GGUF). Attach to any target GGUF to speed up llama.cpp.
poolside Laguna XS 2.1 (33B/3B MoE) quantized by Atomic Chat - GGUF (llama.cpp) and MLX (Apple Silicon) ladders.
-
AtomicChat/Laguna-XS-2.1-GGUF
Text Generation • 33B • Updated • 1.47k • 17 -
AtomicChat/Laguna-XS-2.1-MLX-8bit
Text Generation • 33B • Updated • 512 • 2 -
AtomicChat/Laguna-XS-2.1-MLX-6bit
Text Generation • 33B • Updated • 161 • 2 -
AtomicChat/Laguna-XS-2.1-MLX-5bit
Text Generation • 33B • Updated • 263 • 3
Atomic Chat self-quantized GGUFs of Qwen3.6, built from the original weights with an importance matrix.
Qwen2.5 instruct models, self-quantized to GGUF by Atomic Chat with a per-tensor importance matrix.
Microsoft Phi-4 models, self-quantized to GGUF by Atomic Chat with a per-tensor importance matrix.
DeepReinforce's Ornith 1.0 agentic-coding models, self-quantized to GGUF by Atomic Chat with a per-tensor importance matrix. Runs fully offline.
Atomic Chat self-quantized GGUFs of Qwen3.5, built from the original weights with an importance matrix.
Atomic Chat self-quantized GGUFs of the Qwen3 Coder models, built from the original weights with an importance matrix.
Dynamic-imatrix GGUF quants of Qwen 3.6 27B & 35B-A3B. TurboQuant3 KV + shared-model NextN ready.
Gemma 4 MTP assistant drafters as GGUF (F16/Q8_0/Q5_K_M/Q4_K_M/Q4_K_S). Speculative-decoding heads for the atomic-llama-cpp-turboquant fork.
-
AtomicChat/gemma-4-E2B-it-assistant-GGUF
Text Generation • 77.2M • Updated • 1.29k • 8 -
AtomicChat/gemma-4-E4B-it-assistant-GGUF
Text Generation • 78M • Updated • 1.52k • 26 -
AtomicChat/gemma-4-26B-A4B-it-assistant-GGUF
Text Generation • 0.4B • Updated • 2.58k • 33 -
AtomicChat/gemma-4-31B-it-assistant-GGUF
Text Generation • 0.5B • Updated • 1.95k • 22