Corporate clients are increasingly focusing on AI deployment costs, especially in the context of the continuing increase in large model calls. Writer released a new model and tool to target this pressure point: The implementation costs are kept as low as possible without changing the business usage habits.

Palmyra X6 deployment to enterprise

Writer released a new flagship model, Palmyra X6. The model is based on a post-training version of the Z.ai Open Source Model GLM-5.2, and is positioned for direct use by the enterprise rather than simply seeking benchmark test results.

According to the company, Palmyra X6 would be provided in parallel with other Writer models, as well as external models accessed through Azure or Amazon Bedrock, and the customers would not need to rebuild the whole system around a single model.

Accompanying upgrade targeting token costs

It was published with the new model, and there was a major upgrade of the Writer standard. The focus is not just on the model itself, but rather on the movement, call and reasoning process of the model in carrying out its multi-step tasks.

  • The new model is Palmyra X6
  • Base model from GLM-5.2
  • The cost of basic tasks can be reduced by up to 50 per cent

Writer expects that the new model, with this set of infrastructure adjustments, will result in a 50 per cent reduction in the cost of basic tasks for clients. The company stressed that the overall deployment costs would be significantly reduced if complex tasks were completed at less token and faster.

Research called process optimization more effective

A recent paper by the Writer research team also supports this judgement. The paper tested the cost changes of various models under different implementation processes, and showed that in many cases the optimization of natural efficiency was more stable than a simple replacement model, with an average cost reduction of about 40 per cent.

Writer CEO May Habib stated that the interest of business clients in the continued pursuit of new benchmark tests was diminishing and that the focus was shifting towards the stabilization of costs. She also states that trust in large AI laboratories by technical managers in some enterprises is declining, inter alia, because the latter ' s business model is highly correlated with token usage.