Writer, the full-stack generative AI platform for enterprises, has introduced its latest large language model (LLM), Palmyra X 004. This advanced model aims to revolutionize AI applications and agents by providing the most accurate and reliable means for enterprises to enable actions within generative AI systems. Notably, Palmyra X 004 has achieved the top rank on Berkeley’s Tool Calling Leaderboard, outperforming leading model providers like OpenAI, Anthropic, Meta, and Google.
Key Features of Palmyra X 004
Palmyra X 004 comes with a robust set of features that empower enterprises to harness the full potential of generative AI:
- Actionable AI Capabilities: Enables AI assistants and agents to interact with external systems, performing tasks such as data analysis, code deployment, and workflow execution.
- Tool Calling and API Selection: Achieves nearly a 20% performance edge over competitors on Berkeley’s Tool Calling Leaderboard, showcasing its ability to select the correct tools and execute tasks based on natural language input.
- Automatic Data Integration: Features built-in retrieval augmented generation (RAG) for seamless data integration, allowing for chain-of-thought reasoning and source transparency.
- Expanded Context Window: Offers a 128K context window, enhancing its ability to handle complex interactions and data sets.
May Habib, Writer’s CEO and co-founder, emphasized the importance of these advancements: “Unlocking the full potential of generative AI requires executing complex actions and workflows. Our focus on precision training and architectural innovation positions us to outperform the market while meeting critical enterprise needs.”
Performance Benchmarks
Palmyra X 004 stands out in several evaluations:
- Berkeley’s Tool Calling Leaderboard: With an overall score of 78.76%, it demonstrates exceptional reasoning capabilities in real-world scenarios.
- Stanford University’s HELM Benchmark: It ranks among the top 10 models, scoring 86.1% on HELM Lite and 81.3% on HELM MMLU, which assesses understanding across 57 subjects.
The Future of Generative AI
Writer has a solid track record of model innovation, encompassing open-source, closed environments, and industry-specific models. With the introduction of Palmyra X 004, Writer continues to build on its Palmyra family of enterprise-grade LLMs, which support over 30 languages and multi-modal inputs including images, audio, and video.
Palmyra X 004 is now available for enterprises through various channels, including its API, the Writer Framework, no-code applications, and Slack integration, allowing businesses to seamlessly incorporate advanced AI capabilities into their workflows.
With Palmyra X 004, Writer is setting new standards in generative AI technology. This innovative model not only enhances the functionality of AI applications but also empowers enterprises to streamline their operations, improve decision-making, and drive efficiency in an increasingly competitive landscape.