AI skills and agents for developers

gemma-4-26B-A4B-it Cactus CQ

Cactus-Compute · Kit

2-bit, 3-bit and 4-bit compressed copies of Google's gemma-4-26B-A4B-it chat model in the Cactus weights format. Built to run on ARM devices through the Cactus…

Sarvam-105B GGUF

sarvamai · Kit

The GGUF version of Sarvam-105B, a large mixture-of-experts model with a focus on Indian languages, split into 9 shards at 4-bit for llama.cpp. For people with…

Flint INT4

RecursiveSI · Kit

A 4-bit MLX copy of Recursive Flint for local use on Apple Silicon, about 22.6 GB, launched through Rec with pi or OpenCode. Suits Mac users and developers who…

Qwen2.5-7B-Instruct-AWQ

Qwen · Kit

A 4-bit AWQ build of Qwen2.5-7B-Instruct, a 7.6B chat model that follows instructions, handles tables and returns JSON. For developers who want a capable chat…

Qwen3.8-27B-QUASAR-NVFP4

QUASAR-QAT · Kit

Runs Qwen3.8-27B in 4-bit NVFP4 at 19.7 GB, trained with the QUASAR method to stay close to the full BF16 model. For people serving it with vLLM on an NVIDIA B…

Qwen3.6-35B-A3B-4bit

mlx-community · Kit

A 4-bit MLX conversion of Qwen3.6-35B-A3B that answers text and image prompts on Apple Silicon Macs. Made for people who want to run this Qwen model locally wi…

JevK5-chat-mlx-4bit

mlx-community · Kit

A 4-bit MLX copy of the 4.2B JevK5 chat model for Apple Silicon, with a 262,144 token context and a small memory footprint. Made for Mac users who want a quick…

clef-flash-4bit

mlx-community · Kit

A 4-bit MLX copy of Cloudflare's clef-flash that turns a state and typed questions into a probability for every allowed answer on Apple Silicon Macs. Not a cha…

GLM-5.3-Flash EXL3 4bpw for TensorFold

Mia-AiLab · Kit

4-bit EXL3 quantization of GLM-5.3-Flash prepared for TensorFold, with lower KL divergence than a popular TR3-4bpw quant at the same size. For people serving l…

Granite-4.2-3B (GGUF)

ibm-granite · Kit

Packages IBM's Granite 4.2 3B text model as GGUF files for local use. The 4-bit weights are estimated at about 1.8 GB, for people who want a compact model unde…

gemma-4-31B-it-qat-q4_0-gguf

google · Kit

Runs Google's Gemma 4 31B instruction-tuned model as a 4-bit GGUF file that reads text and images and answers in text. It suits developers and enthusiasts with…

Qwen3.8-27B Splash 8-bit

dmeight · Kit

Runs Qwen3.8-27B at 8-bit through the Splash runtime in LM Studio on Apple silicon. An unofficial package that keeps the model closer to the original than the…

Underdog-Woof-4B-1.1

ConwayResearch · Kit

Generates text as a compact 4.2B model, post-trained from Qwen 3 4B to work better with Underdog's own setup and quantized to 4-bit for MLX. For people running…

Qwen3.5-4B GGUF

bartowski · Kit

Imatrix GGUF quantizations of Qwen3.5-4B, a 4.3B image-text model with MTP layers, with the 4-bit file near 3 GB. For people who run a small model locally.

Llama-3.2-1B-Instruct GGUF

bartowski · Kit

GGUF quantizations of Llama-3.2-1B-Instruct, a 1.2B text model in eight languages, with the 4-bit file near 0.8 GB. For people who run a tiny local model.

Qwen3.8-Whittle-MoE-27B-MLX-4bit

barozp · Kit

A 4-bit MLX build for Apple Silicon of a mixture-of-experts model carved from the dense Qwen3.8-27B, with 17.8B active of 27B total parameters. For people who…

gemma-4-31B-it-qat-GGUF

unsloth · Kit

Runs Google's Gemma 4 31B instruction model from Unsloth's 4-bit QAT GGUF files, with an included drafter for faster speculative decoding in llama.cpp. For peo…

Flint NVFP4 GGUF

RecursiveSI · Kit

A 4-bit NVFP4 copy of the Recursive Flint text model as one GGUF file, built for NVIDIA GPUs through llama.cpp. For developers running a large local model with…

Qwen3.5-0.8B-weight-space-programming-GGUF by NotCherry

NotCherry · Kit

Packages the Qwen3.5-0.8B weight-space-programming model as GGUF files for llama.cpp, with a 4-bit text model and a separate vision projector for image input.

Qwen3.5-9B-OptiQ-4bit

mlx-community · Kit

A 4-bit mixed-precision MLX version of Qwen3.5-9B for local text generation on Apple Silicon, with a bundled multi-token prediction head. For Mac users who run…

GLM-5.3-Flash EXL3 3.5bpw

Mia-AiLab · Kit

Quantized copy of GLM-5.3-Flash in EXL3 format, 156 GB, with narrower routed experts, about 11 percent smaller than the same lab's 4-bit version. For people ru…

K2-Horizon-7B-GGUF

IFM · Kit

Quantized GGUF copies of IFM's K2-Horizon-7B text model for running locally in llama.cpp, from BF16 down to 4-bit, with a 512K-token context window.

Hydrogen-1.0 NVFP4

deep-principle · Kit

A 4-bit NVFP4 build of Hydrogen-1.0, a 381B science and research assistant model from Deep Principle, based on GLM-5.3. It needs a multi-GPU NVIDIA Blackwell s…

Qwen3.8-27B-AWQ-INT4

cyankiwi · Kit

A 4-bit AWQ quantization of Qwen3.8-27B by cyankiwi, a 27B vision-language model with thinking control. Weights are 21.02 GB and load in Transformers, vLLM or…

KAT-Coder-V2.5-Dev GGUF

bartowski · Kit

Imatrix GGUF quantizations of Kwaipilot's KAT-Coder-V2.5-Dev, a 34.7B text model, with 4-bit files of about 19 GB or more. For people who run a large coder-nam…

DeepSeek V4.1 Flash GGUF for DwarfStar

antirez · Kit

Gives Mac users calibrated 2-bit and 4-bit GGUF files of DeepSeek V4.1 Flash, plus a separate vision encoder, for DwarfStar's Metal engine with SSD streaming.…

MNN-Pocket-0.3B-MNN

taobao-mnn · Kit

Runs a 0.3B Chinese and English chat model on phones and edge devices through the MNN runtime, as a 4-bit build of about 173 MB. For developers adding on-devic…

Qwen3.8-27B-Splash

incoai · Kit

Serves Qwen3.8-27B on Apple silicon Macs through Inco AI's Splash engine, with a 4-bit model, a DFlash 2 draft and a vision encoder in one package. Works only…

Qwen3-VL for ComfyUI

Comfy-Org · Kit

Repackaged Qwen3-VL 4B and 8B model files for ComfyUI's text encoder folder, offered in several precisions from bf16 to 4-bit styles. For ComfyUI users who nee…

Wan2.2 TI2V 5B GGUF

QuantStack · Kit

A 5B-parameter text-to-video model converted to GGUF for ComfyUI, about 2.5 GB at 4-bit by estimate, for English and Chinese prompts. It suits ComfyUI users wh…

Qwen3.8-27B-4bit (MLX)

mlx-community · Kit

4-bit MLX conversion of Qwen3.8-27B that reads text and images and answers in text, run with the mlx-vlm tool. For developers who want to run a vision-language…

gemma-4-12B-it-qat-q4_0-gguf

google · Kit

Runs Google's Gemma 4 12B instruction-tuned model in a 4-bit GGUF file made with quantization-aware training. Reads text and images, and suits laptops and cons…

gemma-4-E4B-it-GGUF

unsloth · Kit

A 4-bit GGUF copy of Google's Gemma 4 E4B instruction model, packaged by Unsloth for running on your own machine. It reads text and images and answers in text.

Spark-X2.6-1.7B-GGUF

XHToken · Kit

Runs a small 1.7B on-device language model from XHToken as 4-bit GGUF files, for chat, coding and agent tasks in English and Chinese, with adjustable thinking…

gemma-4-12B-it-qat-GGUF

unsloth · Kit

Gemma 4 12B in GGUF form, trained with quantization-aware training so a 4-bit copy keeps near-bfloat16 quality. For people running Gemma locally with llama.cpp…

Qwen3.8-27B-NVFP4

unsloth · Kit

A 4-bit NVFP4 copy of the Qwen3.8 27B vision-language model made by Unsloth, for serving with vLLM. For developers who want a dense model that reads images and…

Qwen3.8-27B NVFP4

nvidia · Kit

NVIDIA's 4-bit NVFP4 build of Alibaba's Qwen3.8-27B, made with Model Optimizer for Blackwell GPUs, taking text, image and video in and giving text back.

Raon-1 nano (LiteRT-LM)

KRAFTON · Kit

Packs Raon-1 nano as a 4-bit LiteRT-LM bundle with a 4,096-token context, for running a Korean and English chat model on devices. Replies come without a thinki…

Thinklite-27B-GGUF

Sparset · Kit

A 4-bit GGUF copy of Thinklite-27B for llama.cpp that uses less reasoning per answer than its base model, for people running a 27B text model locally.

Raon-1 nano (MLX)

KRAFTON · Kit

Provides Raon-1 nano as 4-bit MLX weights for Apple silicon, so developers can run a Korean and English chat model locally on a Mac. For app builders and local…

dale-0.8b-GGUF

manuelgscom · Kit

A small 4-bit GGUF model that turns a plain request in Spanish or English into one Linux bash command. It runs offline on a CPU inside the dale tool. For peopl…