Gemma 3 is a lightweight open-source multimodal model launched by Google. This version is an instruction-tuned quantization-aware training model with 12B parameters, which has been converted to the MXFP4 format of the MLX framework. It supports text and image input and generates text output, with a 128K context window and support for over 140 languages.

MultimodalMlx

The Brazilian scope routing model is a fine-tuned micro large language model based on Gemma 3 270M, specifically designed for the Brazilian food delivery service system. It can serve as an extremely fast and lightweight intent classifier (router). This model is trained using Unsloth and converted to the GGUF format.

Natural Language ProcessingSafetensorsOther

ZygAI is a fast local large language model designed specifically for Lithuanian and English tasks and is in the testing phase. As a lightweight bilingual assistant, it can run in real-time on local hardware, providing users with efficient and accurate bilingual services.

Natural Language ProcessingGgufMultiple Languages

Lamapi/next-12b is a multilingual large language model with 12 billion parameters, offering multiple quantized versions and supporting various natural language processing tasks such as text generation, question answering, and chatting. This model is trained on multiple domain datasets and features high efficiency and lightweight characteristics.

Natural Language ProcessingTransformersMultiple Languages

The MLX MXFP4 quantized version of Gemma 3 27B IT QAT is a lightweight open - source multimodal model developed by Google. This model can simultaneously process text and image inputs and generate text outputs. It has a large context window of 128K and supports over 140 languages, making it suitable for various text generation and image understanding tasks.

MultimodalMlx

Granite-4.0-H-350M-Base is a lightweight decoder-only language model developed by IBM, designed for resource-constrained devices, supporting multilingual text generation and code completion functions.

Natural Language ProcessingTransformers

This is a sequence-to-sequence model specifically designed to compress short user prompts (≤64 tokens) and can achieve low-latency operation of less than 100 milliseconds on modern GPUs. It serves as a lightweight preprocessing stage to support high-capacity large language models.

Natural Language ProcessingTransformers

NanoAgent is a compact large intelligent language model with 135 million parameters and an 8k context length. It is specifically optimized for tool calls and instruction following, and can run efficiently on personal devices, supporting lightweight inference and ReAct-style interactions.

Natural Language ProcessingMlxEnglish

Granite-4.0-350M-Base is a lightweight decoder-only language model developed by IBM, designed specifically for resource-constrained devices, supporting multilingual text generation and code completion functions.

Natural Language ProcessingTransformers

Granite-4.0-1B-Base is a lightweight decoder-only language model developed by IBM, designed specifically for resource-constrained devices, supporting multilingual text generation and code completion functions.

Natural Language ProcessingTransformers

Granite - 4.0 - H - 350M - Base is a lightweight decoder - only language model developed by IBM, specifically designed for resource - constrained devices. It supports multi - language text generation and code completion functions and can run efficiently on smartphones or IoT devices.

Natural Language ProcessingTransformers

Granite-4.0-H-1B-Base is a lightweight decoder-only language model developed by IBM, designed specifically for resource-constrained devices, supporting multi-language text generation and code completion functions.

Natural Language ProcessingTransformers

KaLM-Embedding-V2.5 is a general-purpose and lightweight embedding model. Through excellent training techniques and data, it achieves the best performance among models of the same scale and can even compete with models 3 - 26 times larger in scale. It supports multiple languages and various embedding dimension options.

Natural Language ProcessingSentence-transformersMultiple Languages

Qwen2.5-14B-Instruct is a large language model with 14 billion parameters, designed specifically for chat and text generation scenarios. This model is built based on the transformers library and is suitable for internal testing and lightweight application deployment.

Natural Language ProcessingTransformersEnglish

GA Guard Lite is a lightweight open-source moderation model developed by General Analysis, specifically designed to detect prohibited content in the outputs of language models. This model can identify seven categories of security risks, including illegal activities, hate speech, and personal information leakage, providing efficient content security protection for AI applications.

Natural Language ProcessingTransformersEnglish

SmolLM-135M-Instruct-manim is a lightweight language model that is fine-tuned via LoRA using the manim code generation dataset based on the HuggingFaceTB/SmolLM-135M-Instruct model. This model is specifically optimized for code generation tasks and offers three different variants to meet various deployment requirements.

Natural Language ProcessingTransformersEnglish

VaultGemma is a lightweight language model pre-trained by Google based on differential privacy technology, with powerful privacy protection capabilities. It uses differential privacy stochastic gradient descent (DP-SGD) for pre-training, providing mathematical privacy guarantees for the training data. It has less than 1 billion parameters and is suitable for a variety of natural language processing tasks.

Natural Language ProcessingTransformers

Jan-v1-AIO-GGUF contains two powerful language models: Jan-v1-4B and Jan-v1-edge. Jan-v1-4B is based on the Qwen3-4B-thinking architecture, with 4 billion parameters. It has strong intelligent reasoning, problem-solving, and tool-using capabilities and supports web search tasks. Jan-v1-edge is a lightweight intelligent model designed for fast execution on devices and is suitable for resource-constrained environments.

Natural Language ProcessingTransformersEnglish

EmbeddingGemma 300M 4bit is a lightweight text embedding model developed by Google, specifically optimized for the MLX framework. This model can convert text into high-quality vector representations and is suitable for various natural language processing tasks, especially sentence similarity calculation and feature extraction.

Natural Language ProcessingSentence-transformers

Gemma 3 270M Instruct is a lightweight instruction-tuned language model developed by Google. Based on a parameter scale of 270M, it is optimized for text generation tasks and supports multiple languages such as Chinese and English.

Natural Language ProcessingTransformers