DeepMath is a mathematical reasoning model with 4 billion parameters, built on Qwen3 - 4B Thinking, combining a fine - tuned large language model and a sandboxed Python executor. It can generate concise Python code snippets for calculation steps, significantly reducing errors and shortening the output length.

Natural Language ProcessingTransformersEnglish

Kimi K2 Thinking is the latest generation of open-source thinking model developed by Moonshot AI. It has powerful deep reasoning ability and tool call function. The model adopts a mixture-of-experts architecture, supports native INT4 quantization, has a 256k context window, and performs excellently in multiple benchmark tests.

Natural Language ProcessingTransformers

DeepSeek-V3.1 is a large language model that is post-trained based on DeepSeek-V3.1-Base. It has a 128K context length and supports mixed thinking modes, intelligent tool invocation, and code proxy functions. This model performs excellently in multiple benchmark tests, especially showing significant improvements in mathematical reasoning, code generation, and search agent tasks.

Natural Language ProcessingTransformersMultiple Languages

DeepSeek-V3.1 is a large language model developed by DeepSeek-AI and is an upgraded version of DeepSeek-V3. This model supports both hybrid thinking mode and non-thinking mode, and performs excellently in tool invocation, code generation, mathematical reasoning, etc., supporting a context length of 128K.

Natural Language ProcessingTransformers

DeepSeek-V3.1 is a large-scale language model developed by the DeepSeek team. It supports both thinking mode and non-thinking mode and performs excellently in multiple benchmark tests. It has powerful text understanding, generation, and reasoning capabilities.

Natural Language ProcessingTransformers

A lightweight language model based on the RWKV-7 architecture, using the flash-linear attention format, with multilingual support and deep thinking ability, suitable for text generation tasks.

Natural Language ProcessingTransformersMultiple Languages

This is a language model specifically optimized for the Indonesian legal field. It is fine-tuned on the Indonesian legal Q&A dataset using the GRPO method based on DeepSeek-R1-0528-Qwen3-8B, focusing on improving legal reasoning and structured thinking abilities.

Natural Language ProcessingTransformers

DeepEyes is a vision-language model that encourages 'thinking with images' through reinforcement learning. It can directly integrate visual information into the reasoning chain and performs excellently in image-text processing tasks.

MultimodalTransformersEnglish

The Deep Thinking Model is a text generation model designed specifically for analytical tasks. It can maintain analytical neutrality when dealing with sensitive queries and ensure the integrity of facts on controversial topics.

Natural Language ProcessingTransformersEnglish

DeepThinkerv2 is a large language model with deep mathematical reasoning capabilities, enhancing problem-solving accuracy through recursive thinking patterns and multi-verification mechanisms.

Natural Language ProcessingSafetensorsEnglish

GLM-4-Z1-32B-0414 is a 32-billion-parameter reasoning model with deep thinking capabilities, excelling in mathematical, coding, and logical tasks

Natural Language ProcessingTransformersMultiple Languages

A truly thinking little mind, employing an extremely deep-narrow structure to capture universal patterns through childlike analogies

Natural Language ProcessingPytorchChinese

LGAI's EXONE series reasoning model, featuring new matrix and extreme quantization technologies, equipped with a 32k context window, focusing on deep thinking and reasoning tasks.

Natural Language ProcessingGgufMultiple Languages

The RWKV-7 g1 model based on the Flash linear attention mechanism, supporting multilingual processing and having deep thinking ability

Natural Language ProcessingTransformersMultiple Languages

DeepHermes is an 8B-parameter language model preview based on the Llama-3 architecture, specializing in deep thinking and long-chain reasoning capabilities.

Natural Language ProcessingGguf

This is an experimentally constructed large language model that combines Dolphin3.0-R1-Mistral-24B with Brainstorm 40x technology, with 45B parameters, and has powerful deep reasoning and thinking abilities. The model supports a 32k context length, uses the CHATML template format, and is specially optimized for detailed output during the reasoning process.

Natural Language ProcessingGgufEnglish

Korean reasoning-enhanced model optimized from DeepSeek-R1-distill-Llama-8B, achieving multilingual output through English internal thinking

Natural Language ProcessingTransformersMultiple Languages

DeepSeek Blackroot is a Deepseek model that integrates a 'thinking/reasoning' distillation component. It is fine-tuned based on Llama and has a certain dark inclination, suitable for both creative and non-creative purposes.

Natural Language ProcessingGgufEnglish