Writer introduces new AI model and upgraded harness to contain token costs
Writer has launched Palmyra X6, a new flagship AI model, alongside upgrades to its agentic “harness” infrastructure, aiming to reduce the cost of enterprise AI deployments. The company says the combined changes could cut customer costs for basic tasks by up to 50%, reflecting growing pressure on businesses to control spending on AI tokens rather than pursue benchmark performance alone.
Palmyra X6 is a post-training adaptation of Z.ai’s open-source GLM-5.2 model and is designed for complex, multi-step work using fewer tokens and completing tasks more quickly. Writer’s research found that improving harness efficiency reduced costs by an average of 40% across tested models, sometimes more reliably than changing the model itself; the new system will be available alongside Writer’s own models and external models accessed through Azure or Amazon Bedrock.
- Writer launched Palmyra X6 and upgraded its AI infrastructure.
- It says basic-task costs could fall by up to 50%.
- The company argues workflow efficiency matters more than model choice.
Both sides, in good faith
The strongest fair case each way — we don't pick a winner.
The case for
Supporters argue that a post-training variant of an open-source base model can give organisations a practical route to capable AI without the expense associated with proprietary frontier systems. They see the upgraded harness as an important engineering contribution: better orchestration, evaluation and cost controls may allow firms to use models more widely while keeping deployments predictable and affordable.
The case against
Sceptics argue that lower token prices alone do not establish that a system is genuinely deployment-ready, particularly for tasks requiring reliability, security and careful handling of sensitive information. They may also question how broadly results based on a particular open-source model will transfer, and contend that customers need transparent independent benchmarks, total operating-cost figures and clear safeguards before treating cost efficiency as a decisive advantage.