Writer launches major agentic AI improvements with Palmyra X6 flagship model

Writer Inc. has unveiled its latest AI model, Palmyra X6, a significant upgrade to the company's flagship generative AI system. This new model aims to provide enterprise-level performance for marketing and revenue teams while maintaining cost efficiency. Alongside the release of Palmyra X6, Writer has also enhanced its AI agent platform, enabling complex, multi-step workflows that prioritize cost-efficiency and long-term task capabilities, even at large scales.

The improvements delivered by Palmyra X6 have resulted in a 52% reduction in cost, a 48% increase in speed, and a 10% boost in quality when compared to Writer's previous flagship model, X5. Co-founder and CTO Waseem AlShikh emphasized the importance of keeping costs flat while allowing enterprises to scale their AI usage, a critical factor for adoption.

Palmyra X6 builds upon the foundation of its predecessor, X5, which was launched in April 2025. The development team focused on refining the model's capabilities to better serve enterprise needs, with a particular emphasis on content generation, sub-agent delegation, and maintaining brand voice consistency. To achieve this, the new model was post-trained on the GLM 5.2, an open-source mixture-of-experts large language model developed by Z.ai Co. Ltd.

During evaluation, Palmyra X6 outperformed other prominent AI models such as Claude Opus 4.8, GPT-5.5, and Gemini 3.1, scoring an average of 0.87 across nine tests. Notably, X6 demonstrated the lowest political bias among the models tested, which is crucial for marketing and revenue teams seeking to produce neutral content that can easily adapt to brand guidelines.

The model completes tasks in an average of 26 seconds, generates 82 tokens per second, and can operate unattended on long-term objectives for up to eight hours without losing focus.

To further streamline the management of these sophisticated AI agents, Writer has introduced enhancements to its reporting and governance backend. Administrators now have a clear overview of activity across teams, with the ability to build and share Playbooks and Skills – repeatable agent workflows and instructions. These Playbooks and Skills can be easily accessed via a dashboard, providing a centralized view of how the AI is being utilized across the business.

By filtering and comparing usage patterns over time, administrators can gain insights into adoption, performance, and expenses, allowing for better decision-making and cost control.

A dedicated section for Playbooks and Skills within the Writer Agent interface offers administrators detailed analytics, monitoring of usage data, managing status and ownership, and tracking performance and costs for individual tasks or skill levels. Additionally, configurable consumption limits and alert thresholds help leadership maintain control over spending across different teams.