Stay updated with our industry-leading AI coverage by subscribing to our daily and weekly newsletters. Learn More
Writer, a leading full-stack generative AI platform, has introduced its latest large language model (LLM) Palmyra X 004 today, marking a significant advancement in enterprise artificial intelligence. This new model excels in function calling and workflow execution, essential capabilities for developing practical AI agents and assistants for businesses.
The launch of Palmyra X 004 comes at a pivotal moment in the AI industry. There is a growing demand for models that can not only process and generate text but also take actions and execute complex workflows as companies strive to integrate generative AI into their operations.
“We are empowering AI to perform multiple functions and actions simultaneously, which is crucial for automating complex enterprise workflows,” stated Waseem Alshikh, Writer’s co-founder and CTO, in an interview with VentureBeat. “With Palmyra X 004, we are transitioning from AI assistants that provide information to systems that can actively work.”
Outperforming tech giants: How Palmyra X 004 is raising the bar for AI function calling
Palmyra X 004 stands out with its exceptional performance on function calling tasks. The model achieved a score of 78.76% on Berkeley’s Tool Calling Leaderboard, surpassing offerings from tech giants like OpenAI, Anthropic, Google, and Meta by nearly 20%. This benchmark assesses a model’s ability to select appropriate tools, determine which APIs to call, and successfully execute tasks based on natural language inputs.
Aside from function calling, Palmyra X 004 also ranked in the top 10 on Stanford University’s Holistic Evaluation of Language Models (HELM) benchmark, scoring 86.1% on HELM Lite and 81.3% on HELM MMLU. These scores indicate strong general language understanding and reasoning abilities across various subjects.
Writer claims to have achieved these results with a model containing only around 150 billion parameters, significantly smaller than some other models with trillions of parameters. The company attributes this efficiency to its innovative use of synthetic data and a proprietary early stopping mechanism during training.
Alshikh explained, “We have developed highly capable models without relying on massive parameter counts or exorbitant training costs. Our model training costs were below a million dollars in GPU time for something above 100 billion parameters. We are demonstrating that you don’t need enormous budgets to compete in the AI space.”
This efficiency-focused approach could have significant implications for the AI industry as companies grapple with the high costs of deploying and running large language models, offering a more affordable and accessible option for enterprise AI solutions.
Breaking barriers: Palmyra X 004’s multilingual and multimodal capabilities
Palmyra X 004 boasts impressive technical specifications, featuring a 128,000 token context window that enables it to process and reason over lengthy documents or conversations. The model supports multilingual capabilities across 30+ languages and can handle multimodal inputs such as text, images, and audio (image and audio capabilities are in beta).
Writer provides multiple deployment options for Palmyra X 004, addressing concerns around data privacy and control. Companies can access the model through Writer’s API, deploy it via cloud providers like AWS SageMaker and Nvidia AI Enterprise, or host the model on-premises within their infrastructure.
The release of Palmyra X 004 mirrors a broader shift in the AI landscape, emphasizing the transformative potential of AI when applied to intricate business processes rather than consumer-facing applications.
“We are witnessing a shift from using AI for simple tasks like email summarization to constructing complex, multi-step workflows,” noted Alshikh. “Our enterprise clients aim to develop AI agents capable of interacting with internal systems, accessing diverse data sources, and executing sophisticated business logic.”
This perspective aligns with industry forecasts, with Gartner predicting that by 2025, 50% of enterprise applications will integrate some form of AI functionality. Writer’s focus on function calling and agentic capabilities positions them well to capitalize on this trend.
The future of AI: Writer’s vision for deeper, smarter, and more efficient models
Despite the progress, challenges persist as AI systems become more deeply embedded in business processes, with issues of reliability, explainability, and governance taking center stage. Writer has taken steps to address some of these concerns with features like automatic data integration with retrieval augmented generation (RAG) and source transparency.
The company emphasizes AI safety and control, with Palmyra X 004 integrating with Writer’s suite of AI guardrails and governance tools, enabling enterprises to establish content policies and manage the model’s outputs.
Looking ahead, Alshikh hinted at Writer’s future research pursuits, exploring the development of even deeper transformer models with potentially 500-2000 layers to enhance reasoning capabilities significantly.
“We are at a pivotal moment in AI advancement,” Alshikh remarked. “The next phase is not just about increasing model size but enhancing intelligence and efficiency. We are concentrating on architectural innovations that can deliver superior reasoning at lower inference costs.”
As the competition in the AI landscape intensifies, Writer’s launch of Palmyra X 004 serves as a testament to the fact that innovation is not solely about scale. By prioritizing efficiency, deployment ease, and real-world business applications, the company is carving a unique path in the enterprise AI sector.
The true measure of success will lie in how enterprises adopt and leverage this technology. As businesses explore the potential of generative AI, models like Palmyra X 004 could play a pivotal role in transforming the promise of AI-driven workflow automation into reality.
