# 32k long context

A code retrieval embedding model trained based on Qwen2, supporting multiple programming languages to facilitate efficient code search and matching.

A Mishima Imatrix quantized version based on Qwen3-4B, enhanced with specific datasets for prose-style generation

Phi-4-reasoning-plus is a large language model developed by Microsoft with enhanced reasoning capabilities, specifically optimized for complex mathematical problems and multi-step reasoning tasks.

A large language model fine-tuned for Norwegian language instructions based on Mistral-7B, developed by the Norwegian University of Science and Technology NowAI Research Center and partners

Cyclone 0.5* is an open-source large language model supporting Russian and English, extending the RoPE context window to 32k and optimizing JSON processing and multi-turn dialogue capabilities

A 70B parameter large language model specialized in dark style creative writing, supporting 32k context length, developed based on miqu-1-70b, excelling in dark/grimdark fantasy themes

Mistral-7B-Instruct-v0.2 is a large language model fine-tuned for instruction following based on Mistral-7B-v0.2, supporting a 32k context window with sliding window attention mechanism removed.