Archive — US — techcrunch
Writer introduces new AI model and upgraded harness to contain token costs
Writer has launched the Palmyra X6 AI model and upgraded harness, aiming to cut deployment costs by up to 50% for marketers. It emphasizes efficiency for complex tasks and was developed from an open-source model, GLM-5.2. CEO May Habib highlights the urgent need for cost reduction in AI deployments, addressing growing enterprise dissatisfaction with major AI labs.
Key points
- New AI model (Palmyra X6) launched by Writer to reduce costs.
- Built on Z.ai’s opensource model GLM-5.2, aimed at providing lower price and deployment-ready capabilities.
- Combination of new model and upgraded infrastructure expected to cut costs by up to 50% for basic AI tasks.
- Focus on multi-step task efficiency, executing faster with fewer tokens.
- Research indicates that harness optimization is a more effective cost-reduction strategy than model choice.
- Model-agnostic deployment allows integration with existing and external models (Azure, Amazon Bedrock).
- CIOs express distrust toward major AI labs due to rising costs and inefficiency in delivering enterprise benefits.
Why it matters
- In 2023, AI cost management is critical as firms seek more efficient solutions amidst rising deployment expenses.
- The shift towards open-source models reflects a broader trend in the industry focused on cost-efficiency.
- The move to optimize AI harnesses aligns with ongoing IT trends prioritizing sustainability and operational efficiency.
Writer's innovation with Palmyra X6 represents a significant step toward addressing cost concerns in AI deployments, reflecting the growing demand for efficiency over mere performance amidst rising operational expenses in the industry.