# WRITER Launches Palmyra X6 and Agent Governance > Source: > Published: 2026-08-14 17:34:20+00:00 # WRITER Launches Palmyra X6 and Agent Governance WRITER released Palmyra X6, a new flagship model, alongside a rebuilt Agent harness and governance tools on August 13. The company reports that the combined model and harness reduce average agent costs by 52%, improve speed by 48%, and raise quality by 10% versus prior performance. The release targets enterprise teams managing token spend across multi-step AI workflows. WRITER released its **Palmyra X6** flagship model, a rebuilt agent orchestration harness, and governance tools on August 13, targeting the cost and operational complexity of production agentic AI. TechCrunch reports that the model and harness became available to WRITER clients the same day. WRITER's model page reports an average finished-task cost of **$0.12**, 52% less than the previous generation. CMSWire, citing the company, reports that WRITER Agent paired with Palmyra X6 runs 48% faster and produces 10% higher-quality results than prior performance. These are vendor-reported figures from WRITER's own evaluations, not independently verified benchmarks. ### Model and harness changes Palmyra X6 is a post-training variation of Z.ai's open-weight GLM-5.2 model, according to TechCrunch and The Next Web. WRITER is using the model as a starting point rather than training Palmyra X6 from scratch. The larger product change is the rebuilt harness, the orchestration layer that manages how an agent executes multi-step work. The Next Web describes the harness as the layer that determines execution across planning, retrieval, tool calls, validation, and retries. In these workflows, token consumption can rise through repeated calls and accumulated context even when the user sees only one final response. TechCrunch reports that WRITER estimates the new model and harness can reduce costs by as much as 50% on basic tasks. It also cites a WRITER research paper finding that harness-efficiency changes reduced costs by an average of 40% across tested models. CMSWire reports that, across the models WRITER tested, the rebuilt harness completed tasks 44% faster and at 41% lower cost on average. WRITER researchers describe the harness as a cross-model efficiency layer: "The harness is the one component whose efficiency multiplies across every model an organization runs—present and future." ### Governance and multi-model operations CMSWire reports that the release adds centralized reporting for adoption, spend, and Playbook and Skill performance. It also reports that administrators can enable Anthropic, OpenAI, or custom models through Microsoft Azure, Amazon Bedrock, or NVIDIA NIM. WRITER's official model page likewise says administrators can bring models from those three platforms. For platform teams, this makes the release relevant beyond a single model benchmark. In comparable enterprise agent deployments, orchestration telemetry and budget controls matter because an agent's cost is shaped by the full execution trace, not solely the input and output price of its selected model. Cross-model observability can also help teams compare routing, retries, context growth, and tool-use patterns under a consistent operating layer. WRITER lists Palmyra X6 pricing at **$2 per million input tokens** and **$8 per million output tokens**, and says an agent can work on one objective for up to eight hours. Enterprises evaluating those claims will need workload-specific testing, particularly for tasks with long context windows, retrieval systems, external tools, and human approval steps, where harness behavior can materially affect both latency and token use. CEO May Habib told TechCrunch, "I think the enterprise is absolutely sick of chasing the next benchmark. They want flattening cost, and it seems like nobody can deliver that." The release places WRITER's public product messaging around operational economics and governance, rather than model capability alone. ## Key Points - 1WRITER released Palmyra X6 with a rebuilt harness, reporting lower task costs, faster execution, and modest quality gains for agent workflows. - 2The release adds centralized spend and adoption reporting, making orchestration telemetry a core operational concern for enterprise AI platform teams. - 3In comparable agent deployments, retries, tool calls, and context growth can dominate total cost beyond a model's advertised token price. ## Scoring Rationale This is a notable enterprise AI platform release focused on a practical production constraint: token costs from multi-step agent execution. Its multi-model harness and governance features are relevant to teams operating agents at scale, although the reported performance gains are vendor claims and not independently verified. ## Sources Primary source and supporting public references used for this report. Practice interview problems based on real data 1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with. [Try 250 free problems](/problems)