Nexusflow Unveils Open-source Generative AI That Empowers Copilots to Use Tools and Outperforms GPT-4

  • NexusRaven-V2 is the latest LLM released by Nexusflow that surpasses GPT-4 in enabling copilots and agents to use software tools
  • This commercially permissive model has the potential to revolutionize workflow automation in complex software, offering a private and affordable solution when deployed at scale

PALO ALTO, Calif.--()--Nexusflow, a generative AI leader targeting the cybersecurity industry, today announced the release of NexusRaven-V2, a 13-billion parameter open-source generative AI model that delivers function calling capability—meaning it can understand human instructions and translate the instructions into precise function/API calls to use a variety of software tools. The function calling capability lies at the core of the OpenAI Assistants API and serves as the key to enabling copilots and agents to use software tools. Instruction-tuned from the CodeLlama-13B model, NexusRaven-V2 achieves up to a 7% higher tool use success rate than the latest OpenAI GPT-4 model on human-curated general software tool use benchmarks.

Nexusflow’s data curation pipeline combines open datasets and Meta’s Llama ecosystem to curate massive high-quality data, which is used to train NexusRaven-V2. The data curation and training of NexusRaven-V2 do not involve any proprietary LLMs such as OpenAI’s GPT-4, which enables enterprise customers to completely own the model that is used to build copilots and agents. This is especially important for enterprise applications which require up-to-date information, quality, safety and in-depth domain customization. NexusRaven-V2, with its superior quality, has the potential to revolutionize workflow automation on complex software with a significantly smaller model size and higher accuracy.

“NexusRaven-V2 outperforms OpenAI’s GPT-4 model head-to-head for using software tools,” said Jiantao Jiao, CEO and co-founder of Nexusflow. “This validates Nexusflow’s technical capability to deliver enterprise solutions using open-source models.”

NexusRaven-V2 demonstrates Nexusflow’s commitment to building effective, reliable generative AI solutions that enable customers’ complete ownership of models and complete control of data. In particular:

  • Nexusflow’s customers can own their cutting-edge generative AI model without data risks associated with proprietary model vendors.
  • Nexusflow’s customers enjoy superior generative AI quality, reliability, cost reduction, and safety with complete and up-to-date customization. This is in comparison to solutions built on proprietary model vendor APIs with unpredictable quality degradation and with lagging knowledge.

Nexusflow is committed to thought leadership for enterprise generative AI solutions and is fully opening its model artifacts to the community to advance the enterprise adoption of generative AI using open-source models.

To learn more about NexusRaven-V2 and its technical capabilities, please visit our blog.

About Nexusflow

Founded by UC Berkeley Computer Science professors and a former SambaNova Systems machine learning director, Nexusflow leverages its deep Generative AI expertise to unlock the full potential of security products and supercharge security operations and engineering. For more information or to get in touch with Nexusflow, visit https://nexusflow.ai/.

Contacts

Liudmila Cherednikova
mila@bospar.com

Release Summary

Nexusflow announces the release of NexusRaven-V2, a 13-billion parameter open-source generative AI model that outperforms GPT-4.

Contacts

Liudmila Cherednikova
mila@bospar.com