tensorblock
Llama-3.2-8B-Instruct-GGUF
llama3.2-1b-Uncensored-GGUF
Phi-4-mini-instruct-abliterated-GGUF
Qwen2.5-7B-Instruct-Uncensored-GGUF
deepseek-coder-7b-instruct-v1.5-GGUF
PocketDoc_Dans-PersonalityEngine-V1.3.0-12b-GGUF
Qwen2.5-32B-Instruct-abliterated-GGUF
mistral-7b-uncensored-GGUF
DarkIdol-Llama-3.1-8B-Instruct-1.2-Uncensored-GGUF
Deepseek-R1-Distill-NSFW-RPv1-GGUF
UnfilteredAI_DAN-Qwen3-1.7B-GGUF
WebSquareAI-Instruct-llama-3-8B-v0.5.39-GGUF
SwallowMaid-8B-L3-SPPO-abliterated-GGUF
llama3.1_korean_v0.1_sft_by_aidx-GGUF
Qwen_Qwen3-Coder-30B-A3B-Instruct-GGUF
BioLlama-Ko-8B-GGUF
Phi-3.5-mini-instruct-GGUF
sexyGPT-Uncensored-GGUF
Qwen2.5-3B-Instruct-GGUF
Midnight-Miqu-70B-v1.5-GGUF
Qwen2.5-3B-Instruct-Uncensored-Test-GGUF
DeepSeek-R1-Distill-Llama-3B-GGUF
NemoMix-Unleashed-12B-GGUF
llama3-8B-slerp-persian-merge-GGUF
Llama-3.2-1B-Instruct-abliterated-GGUF
gemma-2-9b-instruct-GGUF
Meta-Llama-3.1-70B-Instruct-GGUF
Llama3-Aloe-8B-Alpha-GGUF
Llama-3.2-3B-Overthinker-GGUF
Llama-3.1-8B-GGUF
cyber-risk-llama-3-8b-instruct-sft-GGUF
ANIMA-Nectar-v2-GGUF
TowerInstruct-Mistral-7B-v0.2-GGUF
gemma-3-4b-it-GGUF
Xenova_bloom-560m-GGUF
MultiverseEx26-7B-slerp-GGUF
gpt4all-falcon-GGUF
DeepSeek-R1-Qwen2.5-1.5b-SFT-R1-JSON-Unstructured-To-Structured-GGUF
dolphin-2.9.3-llama-3-8b-GGUF
Llama-3-uncensored-Dare-1-GGUF
llama3.2-3b-uncensored-GGUF
Mixtral-8x7B-Instruct-v0.1-GGUF
L500MT-GGUF
L3-8B-Stheno-v3.2-GGUF
ReflectionCoder-DS-33B-GGUF
Calcium-Opus-14B-Elite-1M-GGUF
Yi-6B-200K-GGUF
Qwen2.5-7B-Instruct-GGUF
llama3-eng-ko-8b-sl-GGUF
Llama-3-Instruct-8B-DPO-GGUF
Llama-3.1-8B-Ko-bigdefence-GGUF
tokyotech-llm_Llama-3.1-Swallow-8B-Instruct-v0.5-GGUF
[](https://tensorblock.co) [](https://twitter.com/tensorblockaoi) [](https://discord.gg/Ej5NmeHFf2) [](https://github.com/TensorBlock) [](https://t.me/TensorBlock) tokyotech-llm/Llama-3.1-Swallow-8B-Instruct-v0.5 - GGUF Join our Discord to learn more about what we're building ↗ This repo contains GGUF format model files for tokyotech-llm/Llama-3.1-Swallow-8B-Instruct-v0.5. The files were quantized using machines provided by TensorBlock, and they are compatible with llama.cpp as of commit b5753. A comprehensive collection of Model Context Protocol (MCP) servers. A lightweight, open, and extensible multi-LLM interaction studio. | Filename | Quant type | File Size | Description | | -------- | ---------- | --------- | ----------- | | Llama-3.1-Swallow-8B-Instruct-v0.5-Q2K.gguf | Q2K | 3.179 GB | smallest, significant quality loss - not recommended for most purposes | | Llama-3.1-Swallow-8B-Instruct-v0.5-Q3KS.gguf | Q3KS | 3.665 GB | very small, high quality loss | | Llama-3.1-Swallow-8B-Instruct-v0.5-Q3KM.gguf | Q3KM | 4.019 GB | very small, high quality loss | | Llama-3.1-Swallow-8B-Instruct-v0.5-Q3KL.gguf | Q3KL | 4.322 GB | small, substantial quality loss | | Llama-3.1-Swallow-8B-Instruct-v0.5-Q40.gguf | Q40 | 4.661 GB | legacy; small, very high quality loss - prefer using Q3KM | | Llama-3.1-Swallow-8B-Instruct-v0.5-Q4KS.gguf | Q4KS | 4.693 GB | small, greater quality loss | | Llama-3.1-Swallow-8B-Instruct-v0.5-Q4KM.gguf | Q4KM | 4.921 GB | medium, balanced quality - recommended | | Llama-3.1-Swallow-8B-Instruct-v0.5-Q50.gguf | Q50 | 5.599 GB | legacy; medium, balanced quality - prefer using Q4KM | | Llama-3.1-Swallow-8B-Instruct-v0.5-Q5KS.gguf | Q5KS | 5.599 GB | large, low quality loss - recommended | | Llama-3.1-Swallow-8B-Instruct-v0.5-Q5KM.gguf | Q5KM | 5.733 GB | large, very low quality loss - recommended | | Llama-3.1-Swallow-8B-Instruct-v0.5-Q6K.gguf | Q6K | 6.596 GB | very large, extremely low quality loss | | Llama-3.1-Swallow-8B-Instruct-v0.5-Q80.gguf | Q80 | 8.541 GB | very large, extremely low quality loss - not recommended | Then, downoad the individual model file the a local directory If you wanna download multiple model files with a pattern (e.g., `Q4Kgguf`), you can try:
Llama-3-8B-Lexi-Uncensored-GGUF
llama-3-debug-GGUF
L200MT-GGUF
bloomz-3b-GGUF
gemma-3-12b-it-GGUF
Mistral-Nemo-Instruct-2407-GGUF
WhiteRabbitNeo-2.5-Qwen-2.5-Coder-7B-GGUF
Llama-Primus-Merged-GGUF
AceGPT-v2-8B-Chat-GGUF
llama-3-typhoon-v1.5x-8b-instruct-GGUF
gemma-2-9b-GGUF
Fireball-Alpaca-Llama3.1.08-8B-Philos-C-R1-KTO-beta-GGUF
MagpieLM-8B-SFT-v0.1-GGUF
Qwen2-7B-GGUF
calme-2.6-qwen2-7b-GGUF
Llama-3-Trendyol-LLM-8b-chat-v2.0-GGUF
Lexora-Lite-3B-GGUF
rakeshkiriyath_gpt2Medium_text_to_sql-GGUF
SlimOrca-13B-GGUF
SeaQwen2-0.5B-GGUF
Pinkstack_Base-Roblox-coder-Llama-3.2-3B-vLLM-GGUF
llama2-13b-dpo-v4-GGUF
Qwen1.5-7B-GGUF
DeepSeek-R1-Distill-Llama-8B-abliterated-GGUF
Phi-3-mini-128k-instruct-GGUF
Qra-13B-chat-GGUF
DeepSeek-R1-Strategy-Qwen-2.5-1.5b-Unstructured-To-Structured-GGUF
Configurable-Hermes-3-Llama-3.1-8B-GGUF
Qra-1b-GGUF
granite-8b-code-base-4k-GGUF
llama-3-youko-8b-instruct-GGUF
NemoReRemix-12B-GGUF
tiny-llama3-test-GGUF
Llama-3.1-8B-Lexi-Uncensored-V2-GGUF
DavidAU_Qwen3-8B-64k-Context-2X-Josiefied-Uncensored-GGUF
DeepSeek-V3-1B-Test-GGUF
Llama-3-OpenBioMed-8B-slerp-v0.3-GGUF
SOLAR-10.7B-slerp-GGUF
jpacifico_French-Alpaca-Llama3-8B-Instruct-v1.0-GGUF
starcoder2-15b-instruct-v0.1-GGUF
mistral-7b-grok-GGUF
Llama-3-8B-Instruct-abliterated-v2-GGUF
mims-harvard_TxAgent-T1-Llama-3.1-8B-GGUF
OpenCrystal-12B-L3.1-128K-GGUF
llama3-8B-DarkIdol-2.3-Uncensored-32K-GGUF
SecurityLLM-GGUF
MBZUAI-Paris_Atlas-Chat-2B-GGUF
yodayo-ai_nephra_v1.0-GGUF
salamandra-7b-instruct-GGUF
Llama-medx_v3.1-GGUF
MiniMaxAI_SynLogic-7B-GGUF
[](https://tensorblock.co) [](https://twitter.com/tensorblockaoi) [](https://discord.gg/Ej5NmeHFf2) [](https://github.com/TensorBlock) [](https://t.me/TensorBlock) Join our Discord to learn more about what we're building ↗ This repo contains GGUF format model files for MiniMaxAI/SynLogic-7B. The files were quantized using machines provided by TensorBlock, and they are compatible with llama.cpp as of commit b5753. A comprehensive collection of Model Context Protocol (MCP) servers. A lightweight, open, and extensible multi-LLM interaction studio. | Filename | Quant type | File Size | Description | | -------- | ---------- | --------- | ----------- | | SynLogic-7B-Q2K.gguf | Q2K | 3.016 GB | smallest, significant quality loss - not recommended for most purposes | | SynLogic-7B-Q3KS.gguf | Q3KS | 3.492 GB | very small, high quality loss | | SynLogic-7B-Q3KM.gguf | Q3KM | 3.808 GB | very small, high quality loss | | SynLogic-7B-Q3KL.gguf | Q3KL | 4.088 GB | small, substantial quality loss | | SynLogic-7B-Q40.gguf | Q40 | 4.431 GB | legacy; small, very high quality loss - prefer using Q3KM | | SynLogic-7B-Q4KS.gguf | Q4KS | 4.458 GB | small, greater quality loss | | SynLogic-7B-Q4KM.gguf | Q4KM | 4.683 GB | medium, balanced quality - recommended | | SynLogic-7B-Q50.gguf | Q50 | 5.315 GB | legacy; medium, balanced quality - prefer using Q4KM | | SynLogic-7B-Q5KS.gguf | Q5KS | 5.315 GB | large, low quality loss - recommended | | SynLogic-7B-Q5KM.gguf | Q5KM | 5.445 GB | large, very low quality loss - recommended | | SynLogic-7B-Q6K.gguf | Q6K | 6.254 GB | very large, extremely low quality loss | | SynLogic-7B-Q80.gguf | Q80 | 8.099 GB | very large, extremely low quality loss - not recommended | Then, downoad the individual model file the a local directory If you wanna download multiple model files with a pattern (e.g., `Q4Kgguf`), you can try:
Ko-PlatYi-6B-kiwi-GGUF
defog_sqlcoder2-GGUF
L100MT-GGUF
scb10x_typhoon2.1-gemma3-12b-GGUF
[](https://tensorblock.co) [](https://twitter.com/tensorblockaoi) [](https://discord.gg/Ej5NmeHFf2) [](https://github.com/TensorBlock) [](https://t.me/TensorBlock) Join our Discord to learn more about what we're building ↗ This repo contains GGUF format model files for scb10x/typhoon2.1-gemma3-12b. The files were quantized using machines provided by TensorBlock, and they are compatible with llama.cpp as of commit b5753. A comprehensive collection of Model Context Protocol (MCP) servers. A lightweight, open, and extensible multi-LLM interaction studio. | Filename | Quant type | File Size | Description | | -------- | ---------- | --------- | ----------- | | typhoon2.1-gemma3-12b-Q2K.gguf | Q2K | 4.768 GB | smallest, significant quality loss - not recommended for most purposes | | typhoon2.1-gemma3-12b-Q3KS.gguf | Q3KS | 5.458 GB | very small, high quality loss | | typhoon2.1-gemma3-12b-Q3KM.gguf | Q3KM | 6.009 GB | very small, high quality loss | | typhoon2.1-gemma3-12b-Q3KL.gguf | Q3KL | 6.480 GB | small, substantial quality loss | | typhoon2.1-gemma3-12b-Q40.gguf | Q40 | 6.887 GB | legacy; small, very high quality loss - prefer using Q3KM | | typhoon2.1-gemma3-12b-Q4KS.gguf | Q4KS | 6.935 GB | small, greater quality loss | | typhoon2.1-gemma3-12b-Q4KM.gguf | Q4KM | 7.301 GB | medium, balanced quality - recommended | | typhoon2.1-gemma3-12b-Q50.gguf | Q50 | 8.232 GB | legacy; medium, balanced quality - prefer using Q4KM | | typhoon2.1-gemma3-12b-Q5KS.gguf | Q5KS | 8.232 GB | large, low quality loss - recommended | | typhoon2.1-gemma3-12b-Q5KM.gguf | Q5KM | 8.445 GB | large, very low quality loss - recommended | | typhoon2.1-gemma3-12b-Q6K.gguf | Q6K | 9.661 GB | very large, extremely low quality loss | | typhoon2.1-gemma3-12b-Q80.gguf | Q80 | 12.510 GB | very large, extremely low quality loss - not recommended | Then, downoad the individual model file the a local directory If you wanna download multiple model files with a pattern (e.g., `Q4Kgguf`), you can try:
AMD-Llama-135m-code-GGUF
Qwen2-VL-7B-Instruct-GGUF
dictalm2.0-instruct-GGUF
tabula-8b-GGUF
Aira-2-1B1-GGUF
DeepSeek-R1-Distill-Qwen-7B-abliterated-v2-GGUF
Llama-3.2-1B-Instruct-GGUF
Infinirc-Llama3-8B-2G-Release-v1.0-GGUF
open-llama-3.2-1B-Instruct-GGUF
L3.1-8B-sunfall-stheno-v0.6.1-GGUF
distilgpt2-GGUF
SauerkrautLM-1.5b-GGUF
context_tuned_patient_matching_Llama-3.2-1B-Instruct-GGUF
SOLAR-10.7B-v1.1-GGUF
Python-Code-13B-GGUF
llama3-koen-sft-dpo-v1-GGUF
Meta-Llama-3.1-8B-Instruct-abliterated-GGUF
Colibri_8b_v0.1-GGUF
AMD-Llama-135m-GGUF
LLaMA-Mesh-GGUF
Vikhr-Gemma-2B-instruct-GGUF
granite-20b-code-base-8k-GGUF
QwQ-32B-GGUF
DeepSeek-R1-Distill-Qwen-32B-abliterated-GGUF
MultiLora-drop-sharegpt-GGUF
DCFT-Stratos-Unverified-114k-32B-GGUF
deepseek-r1-14b-cot-math-reasoning-full-GGUF
Llama-3-13B-Instruct-GGUF
gemma-2b-it-GGUF
gpt-neox-20b-GGUF
semi_final_Bllossom-GGUF
Llama-3-70B-Synthia-v3.5-GGUF
zeta-GGUF
BgGPT-Gemma-2-2.6B-IT-v1.0-GGUF
Unbabel_TowerInstruct-13B-v0.1-GGUF
llama-3-sqlcoder-8b-GGUF
llama_16bit_2-GGUF
OpenR1-Qwen-7B-French-GGUF
starcoder2-7b-GGUF
deepseek-coder-1.3b-instruct-GGUF
granite-34b-code-base-8k-GGUF
kyx0r_Neona-12B-GGUF
[](https://tensorblock.co) [](https://twitter.com/tensorblockaoi) [](https://discord.gg/Ej5NmeHFf2) [](https://github.com/TensorBlock) [](https://t.me/TensorBlock) Join our Discord to learn more about what we're building ↗ This repo contains GGUF format model files for kyx0r/Neona-12B. The files were quantized using machines provided by TensorBlock, and they are compatible with llama.cpp as of commit b5753. A comprehensive collection of Model Context Protocol (MCP) servers. A lightweight, open, and extensible multi-LLM interaction studio. | Filename | Quant type | File Size | Description | | -------- | ---------- | --------- | ----------- | | Neona-12B-Q2K.gguf | Q2K | 4.791 GB | smallest, significant quality loss - not recommended for most purposes | | Neona-12B-Q3KS.gguf | Q3KS | 5.534 GB | very small, high quality loss | | Neona-12B-Q3KM.gguf | Q3KM | 6.083 GB | very small, high quality loss | | Neona-12B-Q3KL.gguf | Q3KL | 6.562 GB | small, substantial quality loss | | Neona-12B-Q40.gguf | Q40 | 7.072 GB | legacy; small, very high quality loss - prefer using Q3KM | | Neona-12B-Q4KS.gguf | Q4KS | 7.120 GB | small, greater quality loss | | Neona-12B-Q4KM.gguf | Q4KM | 7.477 GB | medium, balanced quality - recommended | | Neona-12B-Q50.gguf | Q50 | 8.519 GB | legacy; medium, balanced quality - prefer using Q4KM | | Neona-12B-Q5KS.gguf | Q5KS | 8.519 GB | large, low quality loss - recommended | | Neona-12B-Q5KM.gguf | Q5KM | 8.728 GB | large, very low quality loss - recommended | | Neona-12B-Q6K.gguf | Q6K | 10.056 GB | very large, extremely low quality loss | | Neona-12B-Q80.gguf | Q80 | 13.022 GB | very large, extremely low quality loss - not recommended | Then, downoad the individual model file the a local directory If you wanna download multiple model files with a pattern (e.g., `Q4Kgguf`), you can try:
MN-DARKEST-UNIVERSE-29B-GGUF
Monarch-7B-GGUF
Llama-3-Alpha-Ko-8B-Instruct-GGUF
Oolel-v0.1-GGUF
Half-NSFW_Noromaid-7b-GGUF
Qwen-uncensored-v2-GGUF
Sailor-7B-GGUF
Fatgirl_v2_8B-GGUF
FastApply-1.5B-v1.0-GGUF
deepseek-coder-6.7b-instruct-GGUF
SuperNeuralDreadDevil-8b-GGUF
internlm2_5-1_8b-chat-GGUF
Qwen2.5-Coder-32B-Instruct-abliterated-GGUF
SmolLM2-360M-GGUF
Viper-Coder-Hybrid-v1.3-GGUF
Llama-OpenReviewer-8B-GGUF
Mistral-Small3-24B-InstructContinuedFine-GGUF
OLMoE-1B-7B-0924-GGUF
JungZoona_T3Q-qwen2.5-14b-v1.0-e3-GGUF
33x-coder-GGUF
archangel_sft_llama7b-GGUF
AceInstruct-1.5B-GGUF
DeepSeek-R1-Distill-Qwen-1.5B-GGUF
mxbai-rerank-large-v2-GGUF
stablelm-2-12b-GGUF
MunicipalPredictionModel-Llama3-GGUF
huihui-ai_Huihui-Qwen3-4B-abliterated-v2-GGUF
Intelligent-Internet_II-Medical-8B-GGUF
[](https://tensorblock.co) [](https://twitter.com/tensorblockaoi) [](https://discord.gg/Ej5NmeHFf2) [](https://github.com/TensorBlock) [](https://t.me/TensorBlock) Join our Discord to learn more about what we're building ↗ This repo contains GGUF format model files for Intelligent-Internet/II-Medical-8B. The files were quantized using machines provided by TensorBlock, and they are compatible with llama.cpp as of commit b5753. A comprehensive collection of Model Context Protocol (MCP) servers. A lightweight, open, and extensible multi-LLM interaction studio. | Filename | Quant type | File Size | Description | | -------- | ---------- | --------- | ----------- | | II-Medical-8B-Q2K.gguf | Q2K | 3.282 GB | smallest, significant quality loss - not recommended for most purposes | | II-Medical-8B-Q3KS.gguf | Q3KS | 3.770 GB | very small, high quality loss | | II-Medical-8B-Q3KM.gguf | Q3KM | 4.124 GB | very small, high quality loss | | II-Medical-8B-Q3KL.gguf | Q3KL | 4.431 GB | small, substantial quality loss | | II-Medical-8B-Q40.gguf | Q40 | 4.775 GB | legacy; small, very high quality loss - prefer using Q3KM | | II-Medical-8B-Q4KS.gguf | Q4KS | 4.802 GB | small, greater quality loss | | II-Medical-8B-Q4KM.gguf | Q4KM | 5.028 GB | medium, balanced quality - recommended | | II-Medical-8B-Q50.gguf | Q50 | 5.721 GB | legacy; medium, balanced quality - prefer using Q4KM | | II-Medical-8B-Q5KS.gguf | Q5KS | 5.721 GB | large, low quality loss - recommended | | II-Medical-8B-Q5KM.gguf | Q5KM | 5.851 GB | large, very low quality loss - recommended | | II-Medical-8B-Q6K.gguf | Q6K | 6.726 GB | very large, extremely low quality loss | | II-Medical-8B-Q80.gguf | Q80 | 8.710 GB | very large, extremely low quality loss - not recommended | Then, downoad the individual model file the a local directory If you wanna download multiple model files with a pattern (e.g., `Q4Kgguf`), you can try:
gpt2-demo-GGUF
tinyllama-15M-GGUF
zephyr-python-ru-merged-GGUF
Qwen1.5-0.5B-vortex-GGUF
Dolphin3.0-Mistral-24B-GGUF
mamba-2.8b-hf-GGUF
ReadyArt_Broken-Tutu-24B-Unslop-v2.0-GGUF
luckychao_Vicuna-Backdoored-7B-GGUF
Sirius-10B-GGUF
Qwen2.5-Coder-1.5B-GGUF
Aira-2-774M-GGUF
beril-GGUF
Octopus-v2-GGUF
SmolLM2-1.7B-GGUF
llama-3.2-1B-spinquant-hf-GGUF
suzume-llama-3-8B-multilingual-orpo-borda-half-GGUF
Qwen1.5-7B-Chat-GGUF
SherlockAssistant_Mistral-7B-Instruct-Ukrainian-GGUF
Guanaco-3B-Uncensored-v2-GGUF
Mistral-7B-Instruct-v0.2-GGUF
gemma-2-2b-neogenesis-ita-GGUF
Qwen_Qwen3-1.7B-MLX-bf16-GGUF
[](https://tensorblock.co) [](https://twitter.com/tensorblockaoi) [](https://discord.gg/Ej5NmeHFf2) [](https://github.com/TensorBlock) [](https://t.me/TensorBlock) Join our Discord to learn more about what we're building ↗ This repo contains GGUF format model files for Qwen/Qwen3-1.7B-MLX-bf16. The files were quantized using machines provided by TensorBlock, and they are compatible with llama.cpp as of commit b5753. A comprehensive collection of Model Context Protocol (MCP) servers. A lightweight, open, and extensible multi-LLM interaction studio. | Filename | Quant type | File Size | Description | | -------- | ---------- | --------- | ----------- | | Qwen3-1.7B-MLX-bf16-Q2K.gguf | Q2K | 0.778 GB | smallest, significant quality loss - not recommended for most purposes | | Qwen3-1.7B-MLX-bf16-Q3KS.gguf | Q3KS | 0.867 GB | very small, high quality loss | | Qwen3-1.7B-MLX-bf16-Q3KM.gguf | Q3KM | 0.940 GB | very small, high quality loss | | Qwen3-1.7B-MLX-bf16-Q3KL.gguf | Q3KL | 1.004 GB | small, substantial quality loss | | Qwen3-1.7B-MLX-bf16-Q40.gguf | Q40 | 1.054 GB | legacy; small, very high quality loss - prefer using Q3KM | | Qwen3-1.7B-MLX-bf16-Q4KS.gguf | Q4KS | 1.060 GB | small, greater quality loss | | Qwen3-1.7B-MLX-bf16-Q4KM.gguf | Q4KM | 1.107 GB | medium, balanced quality - recommended | | Qwen3-1.7B-MLX-bf16-Q50.gguf | Q50 | 1.231 GB | legacy; medium, balanced quality - prefer using Q4KM | | Qwen3-1.7B-MLX-bf16-Q5KS.gguf | Q5KS | 1.231 GB | large, low quality loss - recommended | | Qwen3-1.7B-MLX-bf16-Q5KM.gguf | Q5KM | 1.258 GB | large, very low quality loss - recommended | | Qwen3-1.7B-MLX-bf16-Q6K.gguf | Q6K | 1.418 GB | very large, extremely low quality loss | | Qwen3-1.7B-MLX-bf16-Q80.gguf | Q80 | 1.834 GB | very large, extremely low quality loss - not recommended | Then, downoad the individual model file the a local directory If you wanna download multiple model files with a pattern (e.g., `Q4Kgguf`), you can try:
chat-gpt2-GGUF
gpt2-GGUF
PiVoT-MoE-GGUF
QwQ-32B-bf16-GGUF
Meta-Llama-3-8B-Instruct-abliterated-v3-GGUF
T-lite-instruct-0.1-GGUF
semcoder_s_1030-GGUF
granite-guardian-3.0-2b-GGUF
Delcos_Mistral-Pygmalion-7b-GGUF
Violet_Twilight-v0.2-GGUF
qwen25-math-7b-instruct-GGUF
calme-2.4-qwen2-7b-GGUF
ross-dev_sexyGPT-Uncensored-GGUF
Phi-4-mini-instruct-GGUF
gemma-ko-1.1-2b-it-GGUF
Llama-3.1-Nemotron-Nano-8B-v1-GGUF
Qwen2.5-Coder-14B-Instruct-abliterated-GGUF
granite-3.1-2b-instruct-GGUF
SWE-Fixer-Retriever-7B-GGUF
defog_sqlcoder-7b-GGUF
kosolra_SFT_DPO_v0-GGUF
s1.1-7B-GGUF
DeepSeek-R1-DRAFT-Qwen2.5-0.5B-GGUF
OpenR1-Qwen-7B-Turkish-GGUF
deepseek-coder-33b-instruct-GGUF
Saul-7B-Instruct-v1-GGUF
SauerkrautLM-Gemma-7b-GGUF
SlimMelodicMaid-GGUF
MFANN-llama3.1-Abliterated-SLERP-GGUF
CohereLabs_aya-23-8B-GGUF
deepseek-math-7b-instruct-GGUF
gemma-3-1b-it-abliterated-GGUF
CHEMLLM-2b-1_5-GGUF
Llama-Song-Stream-3B-Instruct-GGUF
mistral_7b_0-3_oh-dcft-v3.1-claude-3-5-sonnet-20241022-GGUF
ZeroAgency_Mistral-Small-3.1-24B-Instruct-2503-hf-GGUF
ghost-8b-beta-1608-GGUF
Indic-gemma-7b-finetuned-sft-Navarasa-2.0-GGUF
cognitivecomputations_WizardLM-33B-V1.0-Uncensored-GGUF
dolphincoder-starcoder2-15b-GGUF
Triunvirato-7b-GGUF
Qwen2-VL-7B-GGUF
ArliAI_Qwen3-30B-A3B-ArliAI-RpR-v4-Fast-GGUF
DeepSeek-R1-Distill-Qwen-32B-GGUF
macbert4mdcspell_v1-GGUF
mistral-7b-dpo-v6-GGUF
SEOKDONG_llama3.1_korean_v1.1_sft_by_aidx-GGUF
s1k-GGUF
TeenyTinyLlama-160m-GGUF
Yi-34B-200K-DARE-megamerge-v8-GGUF
gemma-2-27b-it-abliterated-GGUF
llm4decompile-6.7b-v2-GGUF
karakuri-lm-8x7b-chat-v0.1-GGUF
Llama-3.1-Instruct_NSFW-pretrained_e1-plus_reddit-GGUF
Mistral-Small-24B-Instruct-2501-abliterated-GGUF
TinyLlama-1.1B-32k-Instruct-GGUF
Tess-2.0-Llama-3-70B-GGUF
mosaicml_mpt-7b-chat-GGUF
Llama-3.1-8B-Ultra-Instruct-GGUF
Lucie-7B-Instruct-v1.1-GGUF
Qwen2.5-Coder-7B-Instruct-GGUF
yamatazen_EtherealAurora-12B-v2-GGUF
mohammedbriman_llama-2-7b-chat-turkish-instructions-GGUF
gpt2-650k-stable-diffusion-prompt-generator-GGUF
Python-Code-33B-GGUF
Llama-3-instruction-constructionsafety-layertuning-GGUF
NSFW_DPO_Noromaid-7b-Mistral-7B-Instruct-v0.1-GGUF
mlx-community_DeepSeek-R1-0528-Qwen3-8B-bf16-GGUF
mrfakename_mistral-small-3.1-24b-base-2503-hf-GGUF
nvidia_AceMath-RL-Nemotron-7B-GGUF
llama-3-8b-gpt-4o-ru1.0-GGUF
chat_gpt2_dpo-GGUF
s1.1-14B-GGUF
prithivMLmods_Ophiuchi-Qwen3-14B-Instruct-GGUF
Roleplay-Llama-3-8B-GGUF
Wayfarer-Large-70B-Llama-3.3-GGUF
DeepSeek-R1-Distill-Qwen-7B-GGUF
DeepSeek-R1-Distill-Llama-70B-abliterated-GGUF
tnayaj-GGUF
h2ogpt-4096-llama2-7b-chat-GGUF
Qwen2-VL-2B-GGUF
TinyLlama-1.1B-Chat-v1.0-GGUF
VityaVitalich_Llama3.1-8b-instruct-GGUF
Q2AW1M-1100-GGUF
shieldgemma-2b-GGUF
Qwen_Qwen3-0.6B-GGUF
gemma-ko-7b-GGUF
shanghong_stage1-GGUF
[](https://tensorblock.co) [](https://twitter.com/tensorblockaoi) [](https://discord.gg/Ej5NmeHFf2) [](https://github.com/TensorBlock) [](https://t.me/TensorBlock) Join our Discord to learn more about what we're building ↗ This repo contains GGUF format model files for shanghong/stage1. The files were quantized using machines provided by TensorBlock, and they are compatible with llama.cpp as of commit b5753. A comprehensive collection of Model Context Protocol (MCP) servers. A lightweight, open, and extensible multi-LLM interaction studio. | Filename | Quant type | File Size | Description | | -------- | ---------- | --------- | ----------- | | stage1-Q2K.gguf | Q2K | 3.179 GB | smallest, significant quality loss - not recommended for most purposes | | stage1-Q3KS.gguf | Q3KS | 3.665 GB | very small, high quality loss | | stage1-Q3KM.gguf | Q3KM | 4.019 GB | very small, high quality loss | | stage1-Q3KL.gguf | Q3KL | 4.322 GB | small, substantial quality loss | | stage1-Q40.gguf | Q40 | 4.661 GB | legacy; small, very high quality loss - prefer using Q3KM | | stage1-Q4KS.gguf | Q4KS | 4.693 GB | small, greater quality loss | | stage1-Q4KM.gguf | Q4KM | 4.921 GB | medium, balanced quality - recommended | | stage1-Q50.gguf | Q50 | 5.599 GB | legacy; medium, balanced quality - prefer using Q4KM | | stage1-Q5KS.gguf | Q5KS | 5.599 GB | large, low quality loss - recommended | | stage1-Q5KM.gguf | Q5KM | 5.733 GB | large, very low quality loss - recommended | | stage1-Q6K.gguf | Q6K | 6.596 GB | very large, extremely low quality loss | | stage1-Q80.gguf | Q80 | 8.541 GB | very large, extremely low quality loss - not recommended | Then, downoad the individual model file the a local directory If you wanna download multiple model files with a pattern (e.g., `Q4Kgguf`), you can try:
llama2-12.8b_lora-dpo_v1-GGUF
YuE-s1-7B-anneal-en-cot-GGUF
sapie1-GGUF
Tifa-Deepsex-14b-CoT-GGUF
Qwen2.5-3B-Instruct-abliterated-GGUF
TinyMistral-6x248M-GGUF
bloom-3b-conversational-GGUF
HelpingAI-3-GGUF
0x-YuAN_codeparrot-ds-GGUF
Llama-3.2-3B-Instruct-uncensored-GGUF
LumiOpen_Llama-Poro-2-8B-Instruct-GGUF
YuE-s1-7B-anneal-zh-icl-GGUF
cotran2_gemma3-1b-GGUF
hivaze_ParaLex-Llama-3-8B-SFT-GGUF
SakanaAI_Llama-3-8B-Instruct-Coding-Expert-GGUF
Themis-GGUF
gpt2023-GGUF
Josiefied-Qwen2.5-7B-Instruct-abliterated-v2-GGUF
pythia-160m-deduped-GGUF
llama2-exams-orca-sharegpt-GGUF
DeepSeek-Coder-V2-Lite-Instruct-GGUF
AetherResearch_Cerebrum-1.0-7b-GGUF
phi-4-GGUF
Qwen1.5-32B-Chat-GGUF
blossom-v3_1-yi-34b-GGUF
gemma-3-1b-it-GGUF
BSC-LT_salamandraTA-2B-GGUF
deepseek-coder-6.7b-base-GGUF
SuperNova-Medius-GGUF
Llama-3.3-70B-Instruct-GGUF
mlx-community_Qwen3-4B-bf16-GGUF
Qwen2.5-Coder-1.5B-Instruct-GGUF
em_german_leo_mistral-GGUF
google_gemma-3-1b-it-GGUF
Phigments12-GGUF
SmolLM2-360M-Instruct-FT-GGUF
DeepSeek-R1-Distill-Qwen-14B-GGUF
Quyen-SE-v0.1-GGUF
Llama-3.2-3B-GGUF
Quble_Test_Model_v1_Pretrain-GGUF
Mixtral-tiny-GGUF
cyber-risk-llama-3-8b-GGUF
OpenLLM-France_Lucie-7B-GGUF
Viking-13B-GGUF
QuyXuan_documents-master-3B-GGUF
[](https://tensorblock.co) [](https://twitter.com/tensorblockaoi) [](https://discord.gg/Ej5NmeHFf2) [](https://github.com/TensorBlock) [](https://t.me/TensorBlock) Join our Discord to learn more about what we're building ↗ This repo contains GGUF format model files for QuyXuan/documents-master-3B. The files were quantized using machines provided by TensorBlock, and they are compatible with llama.cpp as of commit b5753. A comprehensive collection of Model Context Protocol (MCP) servers. A lightweight, open, and extensible multi-LLM interaction studio. | Filename | Quant type | File Size | Description | | -------- | ---------- | --------- | ----------- | | documents-master-3B-Q2K.gguf | Q2K | 1.364 GB | smallest, significant quality loss - not recommended for most purposes | | documents-master-3B-Q3KS.gguf | Q3KS | 1.543 GB | very small, high quality loss | | documents-master-3B-Q3KM.gguf | Q3KM | 1.687 GB | very small, high quality loss | | documents-master-3B-Q3KL.gguf | Q3KL | 1.815 GB | small, substantial quality loss | | documents-master-3B-Q40.gguf | Q40 | 1.917 GB | legacy; small, very high quality loss - prefer using Q3KM | | documents-master-3B-Q4KS.gguf | Q4KS | 1.928 GB | small, greater quality loss | | documents-master-3B-Q4KM.gguf | Q4KM | 2.019 GB | medium, balanced quality - recommended | | documents-master-3B-Q50.gguf | Q50 | 2.270 GB | legacy; medium, balanced quality - prefer using Q4KM | | documents-master-3B-Q5KS.gguf | Q5KS | 2.270 GB | large, low quality loss - recommended | | documents-master-3B-Q5KM.gguf | Q5KM | 2.322 GB | large, very low quality loss - recommended | | documents-master-3B-Q6K.gguf | Q6K | 2.644 GB | very large, extremely low quality loss | | documents-master-3B-Q80.gguf | Q80 | 3.422 GB | very large, extremely low quality loss - not recommended | Then, downoad the individual model file the a local directory If you wanna download multiple model files with a pattern (e.g., `Q4Kgguf`), you can try:
calme-3.1-qwenloi-3b-GGUF
gemma-2-2b-it-abliterated-GGUF
Qwen1.5-MoE-A2.7B-Chat-GGUF
Meta-Llama-3-8B-hf-GGUF
Blue-Orchid-2x7b-GGUF
SeaLLM-7B-v2.5-GGUF
Arcee-Spark-GGUF
Velara-11B-V2-GGUF
ghost-8b-beta-GGUF
occiglot-7b-it-en-instruct-GGUF
Llama-3.1-Swallow-8B-Instruct-v0.3-GGUF
anime-anything-promptgen-v2-GGUF
Qwen1.5-1.8B-GGUF
Llama-3.2-1B-GGUF
Goekdeniz-Guelmez_Josiefied-DeepSeek-R1-0528-Qwen3-8B-abliterated-v1-GGUF
s1-0.5B-GGUF
Qwen2.5-Coder-7B-GGUF
Josiefied-Qwen2.5-14B-Instruct-abliterated-v4-GGUF
llama-3-2-1b-sft-GGUF
TenyxChat-7B-v1-GGUF
gpt2-medium-GGUF
agentica-org_DeepSWE-Preview-GGUF
shuttleai_shuttle-3.5-GGUF
Meltemi-7B-Instruct-v1.5-GGUF
futurehouse_ether0-GGUF
Ambari-7B-Instruct-v0.1-sharded-GGUF
FreedomIntelligence_HuatuoGPT-Vision-7B-Qwen2.5VL-GGUF
Qwen1.5-32B-GGUF
MiniCPM-2B-128k-GGUF
gemma2-gutenberg-27B-GGUF
Qwen2.5-Coder-32B-Instruct-GGUF
ChimeraLlama-3-8B-v2-GGUF
tnayajv2.0-GGUF
starcoder2-3b-GGUF
saiga_llama3_8b-GGUF
SauerkrautLM-Qwen-32b-GGUF
vinallama-2.7b-chat-GGUF
orca_mini_3b-GGUF
TarhanE_GRPO-Qwen3-0.6B-GGUF
[](https://tensorblock.co) [](https://twitter.com/tensorblockaoi) [](https://discord.gg/Ej5NmeHFf2) [](https://github.com/TensorBlock) [](https://t.me/TensorBlock) Join our Discord to learn more about what we're building ↗ This repo contains GGUF format model files for TarhanE/GRPO-Qwen3-0.6B. The files were quantized using machines provided by TensorBlock, and they are compatible with llama.cpp as of commit b5753. A comprehensive collection of Model Context Protocol (MCP) servers. A lightweight, open, and extensible multi-LLM interaction studio. | Filename | Quant type | File Size | Description | | -------- | ---------- | --------- | ----------- | | GRPO-Qwen3-0.6B-Q2K.gguf | Q2K | 0.296 GB | smallest, significant quality loss - not recommended for most purposes | | GRPO-Qwen3-0.6B-Q3KS.gguf | Q3KS | 0.323 GB | very small, high quality loss | | GRPO-Qwen3-0.6B-Q3KM.gguf | Q3KM | 0.347 GB | very small, high quality loss | | GRPO-Qwen3-0.6B-Q3KL.gguf | Q3KL | 0.368 GB | small, substantial quality loss | | GRPO-Qwen3-0.6B-Q40.gguf | Q40 | 0.382 GB | legacy; small, very high quality loss - prefer using Q3KM | | GRPO-Qwen3-0.6B-Q4KS.gguf | Q4KS | 0.383 GB | small, greater quality loss | | GRPO-Qwen3-0.6B-Q4KM.gguf | Q4KM | 0.397 GB | medium, balanced quality - recommended | | GRPO-Qwen3-0.6B-Q50.gguf | Q50 | 0.437 GB | legacy; medium, balanced quality - prefer using Q4KM | | GRPO-Qwen3-0.6B-Q5KS.gguf | Q5KS | 0.437 GB | large, low quality loss - recommended | | GRPO-Qwen3-0.6B-Q5KM.gguf | Q5KM | 0.444 GB | large, very low quality loss - recommended | | GRPO-Qwen3-0.6B-Q6K.gguf | Q6K | 0.495 GB | very large, extremely low quality loss | | GRPO-Qwen3-0.6B-Q80.gguf | Q80 | 0.639 GB | very large, extremely low quality loss - not recommended | Then, downoad the individual model file the a local directory If you wanna download multiple model files with a pattern (e.g., `Q4Kgguf`), you can try:
Machina_24B.V2-GGUF
reka-flash-3-GGUF
L3.2-Rogue-Creative-Instruct-Uncensored-Abliterated-7B-GGUF
Unbabel_Tower-Plus-9B-GGUF
[](https://tensorblock.co) [](https://twitter.com/tensorblockaoi) [](https://discord.gg/Ej5NmeHFf2) [](https://github.com/TensorBlock) [](https://t.me/TensorBlock) Join our Discord to learn more about what we're building ↗ This repo contains GGUF format model files for Unbabel/Tower-Plus-9B. The files were quantized using machines provided by TensorBlock, and they are compatible with llama.cpp as of commit b5753. A comprehensive collection of Model Context Protocol (MCP) servers. A lightweight, open, and extensible multi-LLM interaction studio. | Filename | Quant type | File Size | Description | | -------- | ---------- | --------- | ----------- | | Tower-Plus-9B-Q2K.gguf | Q2K | 3.805 GB | smallest, significant quality loss - not recommended for most purposes | | Tower-Plus-9B-Q3KS.gguf | Q3KS | 4.338 GB | very small, high quality loss | | Tower-Plus-9B-Q3KM.gguf | Q3KM | 4.762 GB | very small, high quality loss | | Tower-Plus-9B-Q3KL.gguf | Q3KL | 5.132 GB | small, substantial quality loss | | Tower-Plus-9B-Q40.gguf | Q40 | 5.443 GB | legacy; small, very high quality loss - prefer using Q3KM | | Tower-Plus-9B-Q4KS.gguf | Q4KS | 5.479 GB | small, greater quality loss | | Tower-Plus-9B-Q4KM.gguf | Q4KM | 5.761 GB | medium, balanced quality - recommended | | Tower-Plus-9B-Q50.gguf | Q50 | 6.484 GB | legacy; medium, balanced quality - prefer using Q4KM | | Tower-Plus-9B-Q5KS.gguf | Q5KS | 6.484 GB | large, low quality loss - recommended | | Tower-Plus-9B-Q5KM.gguf | Q5KM | 6.647 GB | large, very low quality loss - recommended | | Tower-Plus-9B-Q6K.gguf | Q6K | 7.589 GB | very large, extremely low quality loss | | Tower-Plus-9B-Q80.gguf | Q80 | 9.827 GB | very large, extremely low quality loss - not recommended | Then, downoad the individual model file the a local directory If you wanna download multiple model files with a pattern (e.g., `Q4Kgguf`), you can try:
Viking-33B-GGUF
MBZUAI-Paris_Nile-Chat-4B-GGUF
Breeze-7B-Instruct-v1_0-GGUF
Light-R1-32B-DS-GGUF
OpenHermes-2.5-Mistral-7B-pruned50-GGUF
DeepSeek-R1-Distill-Qwen-Coder-32B-Fusion-9010-GGUF
cognitivecomputations_samantha-mistral-instruct-7b-GGUF
mt0-xxl-mt-GGUF
swordfaith_ReTool-Qwen3-4B-SFT-cold-started-GGUF
[](https://tensorblock.co) [](https://twitter.com/tensorblockaoi) [](https://discord.gg/Ej5NmeHFf2) [](https://github.com/TensorBlock) [](https://t.me/TensorBlock) Join our Discord to learn more about what we're building ↗ This repo contains GGUF format model files for swordfaith/ReTool-Qwen3-4B-SFT-cold-started. The files were quantized using machines provided by TensorBlock, and they are compatible with llama.cpp as of commit b5753. A comprehensive collection of Model Context Protocol (MCP) servers. A lightweight, open, and extensible multi-LLM interaction studio. | Filename | Quant type | File Size | Description | | -------- | ---------- | --------- | ----------- | | ReTool-Qwen3-4B-SFT-cold-started-Q2K.gguf | Q2K | 1.669 GB | smallest, significant quality loss - not recommended for most purposes | | ReTool-Qwen3-4B-SFT-cold-started-Q3KS.gguf | Q3KS | 1.887 GB | very small, high quality loss | | ReTool-Qwen3-4B-SFT-cold-started-Q3KM.gguf | Q3KM | 2.076 GB | very small, high quality loss | | ReTool-Qwen3-4B-SFT-cold-started-Q3KL.gguf | Q3KL | 2.240 GB | small, substantial quality loss | | ReTool-Qwen3-4B-SFT-cold-started-Q40.gguf | Q40 | 2.370 GB | legacy; small, very high quality loss - prefer using Q3KM | | ReTool-Qwen3-4B-SFT-cold-started-Q4KS.gguf | Q4KS | 2.383 GB | small, greater quality loss | | ReTool-Qwen3-4B-SFT-cold-started-Q4KM.gguf | Q4KM | 2.497 GB | medium, balanced quality - recommended | | ReTool-Qwen3-4B-SFT-cold-started-Q50.gguf | Q50 | 2.824 GB | legacy; medium, balanced quality - prefer using Q4KM | | ReTool-Qwen3-4B-SFT-cold-started-Q5KS.gguf | Q5KS | 2.824 GB | large, low quality loss - recommended | | ReTool-Qwen3-4B-SFT-cold-started-Q5KM.gguf | Q5KM | 2.890 GB | large, very low quality loss - recommended | | ReTool-Qwen3-4B-SFT-cold-started-Q6K.gguf | Q6K | 3.306 GB | very large, extremely low quality loss | | ReTool-Qwen3-4B-SFT-cold-started-Q80.gguf | Q80 | 4.280 GB | very large, extremely low quality loss - not recommended | Then, downoad the individual model file the a local directory If you wanna download multiple model files with a pattern (e.g., `Q4Kgguf`), you can try:
Mistral_solar-slerp-GGUF
Uncensored_llama_3.2_3b_safetensors-GGUF
aya-expanse-8b-GGUF
Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss-GGUF
X-ALMA-13B-Pretrain-GGUF
tempesthenno-nuslerp-0124-GGUF
zephyr-7b-beta-GGUF
deepseek-coder-7b-base-v1.5-GGUF
ghost-7b-v0.9.0-GGUF
MARS-GGUF
MNCJihunKim_Mistral-7B-SlimOrca-OP-8k-GGUF
MediKAI-GGUF
llama3.1_1B_adapted-GGUF
Jarvis1111_DoctorAgent-RL-SFT-1k-Thinking-GGUF
Qwen2-VL-2B-Instruct-GGUF
FractalAIResearch_Fathom-R1-14B-GGUF
castorini_rank_vicuna_7b_v1_fp16-GGUF
llama3-diverce-ver1.0-GGUF
Llama-3.1-8B-Lexi-Uncensored-GGUF
HuatuoGPT-o1-72B-GGUF
LLaMA3-iterative-DPO-final-ExPO-GGUF
llama3_generative_qa_2-GGUF
Qwen_Qwen3-8B-MLX-bf16-GGUF
kakaocorp_kanana-1.5-2.1b-instruct-2505-GGUF
llama3-math-trans-sft-GGUF
Qwen_Qwen3-1.7B-GGUF
Hermes-3-Llama-3.1-70B-GGUF
Llama-3-WhiteRabbitNeo-8B-v2.0-GGUF
Smaug-34B-v0.1-GGUF
Sailor2-20B-Chat-GGUF
CodeLlama-70b-Python-hf-GGUF
TowerBase-7B-v0.1-GGUF
Qwen2.5-Coder-32B-GGUF
TheBloke_Wizard-Vicuna-30B-Uncensored-fp16-GGUF
WizardLM-30B-Uncensored-Guanaco-SuperCOT-30b-GGUF
mixtralnt-4x7b-test-GGUF
RankingGPT-bloom-560m-GGUF
dolphin-2.9.1-llama-3-8b-GGUF
Llama-3-ELYZA-JP-8B-GGUF
granite-7b-instruct-GGUF
granite-8b-code-instruct-4k-GGUF
cybersentinal-2.0-GGUF
CohereLabs_c4ai-command-r-08-2024-GGUF
c4ai-command-r-v01-GGUF
mlabonne_gemma-3-12b-it-qat-abliterated-GGUF
[](https://tensorblock.co) [](https://twitter.com/tensorblockaoi) [](https://discord.gg/Ej5NmeHFf2) [](https://github.com/TensorBlock) [](https://t.me/TensorBlock) Join our Discord to learn more about what we're building ↗ This repo contains GGUF format model files for mlabonne/gemma-3-12b-it-qat-abliterated. The files were quantized using machines provided by TensorBlock, and they are compatible with llama.cpp as of commit b5753. A comprehensive collection of Model Context Protocol (MCP) servers. A lightweight, open, and extensible multi-LLM interaction studio. | Filename | Quant type | File Size | Description | | -------- | ---------- | --------- | ----------- | | gemma-3-12b-it-qat-abliterated-Q2K.gguf | Q2K | 4.768 GB | smallest, significant quality loss - not recommended for most purposes | | gemma-3-12b-it-qat-abliterated-Q3KS.gguf | Q3KS | 5.458 GB | very small, high quality loss | | gemma-3-12b-it-qat-abliterated-Q3KM.gguf | Q3KM | 6.009 GB | very small, high quality loss | | gemma-3-12b-it-qat-abliterated-Q3KL.gguf | Q3KL | 6.480 GB | small, substantial quality loss | | gemma-3-12b-it-qat-abliterated-Q40.gguf | Q40 | 6.887 GB | legacy; small, very high quality loss - prefer using Q3KM | | gemma-3-12b-it-qat-abliterated-Q4KS.gguf | Q4KS | 6.935 GB | small, greater quality loss | | gemma-3-12b-it-qat-abliterated-Q4KM.gguf | Q4KM | 7.301 GB | medium, balanced quality - recommended | | gemma-3-12b-it-qat-abliterated-Q50.gguf | Q50 | 8.232 GB | legacy; medium, balanced quality - prefer using Q4KM | | gemma-3-12b-it-qat-abliterated-Q5KS.gguf | Q5KS | 8.232 GB | large, low quality loss - recommended | | gemma-3-12b-it-qat-abliterated-Q5KM.gguf | Q5KM | 8.445 GB | large, very low quality loss - recommended | | gemma-3-12b-it-qat-abliterated-Q6K.gguf | Q6K | 9.661 GB | very large, extremely low quality loss | | gemma-3-12b-it-qat-abliterated-Q80.gguf | Q80 | 12.510 GB | very large, extremely low quality loss - not recommended | Then, downoad the individual model file the a local directory If you wanna download multiple model files with a pattern (e.g., `Q4Kgguf`), you can try:
llama3-8B-DarkIdol-2.2-Uncensored-1048K-GGUF
stablelm-3b-4e1t-GGUF
marin-community_marin-8b-base-GGUF
JSL-MedLlama-3-8B-v1.0-GGUF
Qwen2.5-1.5B-Instruct-GGUF
Qwen2.5-Coder-14B-Instruct-GGUF
granite-3.0-8b-instruct-GGUF
Qwen1.5-MoE-A2.7B-GGUF
Salesforce_xgen-small-9B-instruct-r-GGUF
salamandra-2b-instruct-GGUF
SOLAR-10.7B-v1.0-GGUF
redrix_patricide-12B-Unslop-Mell-GGUF
JSL-MedLlama-3-8B-v2.0-GGUF
SSH_355M-GGUF
TinyLlama-1.1B-intermediate-step-1195k-token-2.5T-GGUF
Yi-Ko-6B-Instruct-v1.0-GGUF
codellama-13b-instruct-nf4-fp16-upscaled-GGUF
L3H10M-0000-GGUF
MLlamav1-GGUF
ONS-SOLAR-10.7B-v1.2-GGUF
open-llama-3b-v2-elmv3-GGUF
Llama-3.2-3B-Instruct-GGUF
llama-3-8b-GGUF
WiroAI-Finance-Qwen-1.5B-GGUF
sft-ds-140k-GGUF
saiga_tlite_8b-GGUF
Amal-70b-GGUF
seeklhy_codes-7b-spider-GGUF
r1-1776-GGUF
Qwen2.5-14B-Instruct-GGUF
Meta-Llama-3.1-8B-Instruct-GGUF
andresnowak_Qwen3-0.6B-instruction-finetuned-GGUF
Samantha2.0-Phi4-ita-16bit-GGUF
bloom-1b1-GGUF
llama2-13B-eugeneparkthebest-GGUF
Felladrin_TinyMistral-248M-Chat-v2-GGUF
L3.1-Suze-Vume-2-calc-GGUF
s1K_32b-GGUF
Qwen_Qwen3-8B-GGUF
unsloth_Qwen3-30B-A3B-Instruct-2507-GGUF
Qwen2.5-7B-nerd-uncensored-v1.0-GGUF
Qwen_Qwen3-4B-GGUF
redpajama-3b-chat-GGUF
CodeLlama-34b-Python-hf-GGUF
llama-3-8b-Instruct-GGUF
LLaMA3-SFT-v2-GGUF
Llama-3-8B-Stroganoff-GGUF
GritLM-8x7B-GGUF
MiniPLM-Qwen-200M-GGUF
Mistral-7B-Merge-14-v0.3-GGUF
Orion-14B-Base-GGUF
FusionNet-GGUF
falcon-7b-instruct-GGUF
Aspik101_Vicuzard-30B-Uncensored-instruct-PL-lora_unload-GGUF
PhysicsWallahAI_Aryabhata-1.0-GGUF
[](https://tensorblock.co) [](https://twitter.com/tensorblockaoi) [](https://discord.gg/Ej5NmeHFf2) [](https://github.com/TensorBlock) [](https://t.me/TensorBlock) Join our Discord to learn more about what we're building ↗ This repo contains GGUF format model files for PhysicsWallahAI/Aryabhata-1.0. The files were quantized using machines provided by TensorBlock, and they are compatible with llama.cpp as of commit b5753. A comprehensive collection of Model Context Protocol (MCP) servers. A lightweight, open, and extensible multi-LLM interaction studio. | Filename | Quant type | File Size | Description | | -------- | ---------- | --------- | ----------- | | Aryabhata-1.0-Q2K.gguf | Q2K | 3.016 GB | smallest, significant quality loss - not recommended for most purposes | | Aryabhata-1.0-Q3KS.gguf | Q3KS | 3.492 GB | very small, high quality loss | | Aryabhata-1.0-Q3KM.gguf | Q3KM | 3.808 GB | very small, high quality loss | | Aryabhata-1.0-Q3KL.gguf | Q3KL | 4.088 GB | small, substantial quality loss | | Aryabhata-1.0-Q40.gguf | Q40 | 4.431 GB | legacy; small, very high quality loss - prefer using Q3KM | | Aryabhata-1.0-Q4KS.gguf | Q4KS | 4.458 GB | small, greater quality loss | | Aryabhata-1.0-Q4KM.gguf | Q4KM | 4.683 GB | medium, balanced quality - recommended | | Aryabhata-1.0-Q50.gguf | Q50 | 5.315 GB | legacy; medium, balanced quality - prefer using Q4KM | | Aryabhata-1.0-Q5KS.gguf | Q5KS | 5.315 GB | large, low quality loss - recommended | | Aryabhata-1.0-Q5KM.gguf | Q5KM | 5.445 GB | large, very low quality loss - recommended | | Aryabhata-1.0-Q6K.gguf | Q6K | 6.254 GB | very large, extremely low quality loss | | Aryabhata-1.0-Q80.gguf | Q80 | 8.099 GB | very large, extremely low quality loss - not recommended | Then, downoad the individual model file the a local directory If you wanna download multiple model files with a pattern (e.g., `Q4Kgguf`), you can try: