{"type":"rich","version":"1.0","provider_name":"Transistor","provider_url":"https://transistor.fm","author_name":"AI News Today | Julian Goldie Podcast","title":"New Chinese AI Model Is INSANE! (FREE & Open Source)","html":"<iframe width=\"100%\" height=\"180\" frameborder=\"no\" scrolling=\"no\" seamless src=\"https://share.transistor.fm/e/04dbc23d\"></iframe>","width":"100%","height":180,"duration":319,"description":"N2 Pro: Free Open-Source Chinese Coding Model Beating Claude & GPT on Benchmarks (260K Context)\nThe video introduces N2, a new free Chinese open-source model with open weights and a 260K-token context window, designed for coding and agentic workflows with strong tool/function calling. It highlights benchmark results where N2 outperforms paid APIs like Claude Opus 4.7, GPT 5.5, and DeepSeek V4 Pro on some tests (including SWE Bench), and notes it is based on Qwen 3.5’s architecture. The model is available on Hugging Face and via a free API on OpenRouter, and can be used inside agents such as Kilocode, Hermes Agent, Claude Code (via an open-source “Free Claude Code” project), OpenClaude, and Pi. The creator demonstrates configuring N2 in Hermes Agent using model profiles, compares N2 Pro and N2 Mini, and shows N2 responding faster than a Gemma 4 setup. The video ends by promoting training and resources in the AI Profit Boardroom.\n00:00 Free N2 Model Reveal\n00:29 Benchmarks and Specs\n00:58 Where to Get It Free\n01:24 Setup in Hermes Agent\n02:01 Use with Claude Code\n02:16 What You Can Build\n02:32 DeepSeek Comparison\n02:58 Use Cases and Variants\n03:28 Speed Test vs Gemma\n04:18 Wrap Up and Next Steps\n04:30 AI Profit Boardroom Pitch\n05:17 Final Thanks","thumbnail_url":"https://img.transistorcdn.com/Tp0JSUlyiAe2ran44b13cjk4ImQYS1QEMKCAa9Q7go0/rs:fill:0:0:1/w:400/h:400/q:60/mb:500000/aHR0cHM6Ly9pbWct/dXBsb2FkLXByb2R1/Y3Rpb24udHJhbnNp/c3Rvci5mbS9lNTYz/NGQxMTU2ZTM1MzU5/MDNiYjcwZjJmYTY5/ODJmOS5qcGc.webp","thumbnail_width":300,"thumbnail_height":300}