Menlo/Jan-nano

Jan-Nano is a compact 4-billion parameter language model developed by Menlo Research, specifically designed and trained for deep research tasks. It is optimized to work seamlessly with Model Context Protocol (MCP) servers, enabling efficient integration with various research tools and data sources. With a context length of 40960 tokens, Jan-Nano excels in factual accuracy and effectiveness within MCP-enabled environments, making it suitable for tool-augmented research. It demonstrates strong performance on the SimpleQA benchmark for its size.