Katanemo’s Arch-Function LLMs

• 

  Katanemo’s Arch-Function LLMs: Fast and Cost-Effective Workflows

Generative AI is advancing rapidly, enabling businesses to use AI agents for digital tasks, but many struggle with slow performance and high costs. Katanemo, a startup, has introduced Arch-Function, an open-sourcing ultra-fast large language model (LLM) 12 times faster than OpenAI’s GPT-4 and more cost-effective. These models excel in function-calling tasks, and streamlining workflows by interacting with external systems efficiently. Built on Qwen 2.5 with 3B and 7B parameters, they outperform competitors in speed, cost, and reliability. This innovation paves the way for smarter, faster, and budget-friendly AI applications, transforming enterprise software by 2028.

Built With

• canva
• promptspad