OpenAI o1-like CoT and logical thinking prompting strategy significantly enhances llm responses.

PixelVerse t1 is powered by Llama 3.1 70b and 3.2 90b. However, with a detailed CoT prompt, it answers complex questions correctly, much better than it's base model and even sometimes beating flagship models like GPT 4o, Claude 3 and Gemini.

Try it at: https://ai.pixelverse.tech/app/cortexchat