Mentat serves as a cryptocurrency intelligence platform designed to enhance Web3 agentic frameworks and consumer applications. It enables both users and applications to seamlessly access blockchain information, facilitating a wide range of tasks from simple insights to intricate research and detailed multi-step analyses. The platform merges agentic planning with extensive access to DeFi protocols, market insights, and cross-chain functionalities, allowing it to autonomously process large quantities of real-time data. Mentat can manage everything from specific activities like analyzing transaction records and assessing token distributions, to more sophisticated operations such as conducting liquidity pool evaluations. Looking ahead, we plan to introduce capabilities for more advanced on-chain tasks, including identifying security risks in smart contracts, carrying out complex DeFi transactions, and orchestrating and refining on-chain governance proposals. This creates a remarkably robust system that makes advanced blockchain comprehension accessible to everyone, from casual investors to developers, enabling data-driven decision-making while sidestepping intricate technical challenges. The innovation we are rolling out is based on a new generation of AI systems commonly known as reasoning models. These models, developed by leading research institutions and now increasingly utilized across the AI landscape, have the capability to interpret text commands, convert them into actionable steps, and intelligently link various tools to execute intricate tasks. They go beyond basic "question-and-answer" exchanges and are capable of managing entire workflows. This transition marks a significant advancement from traditional language models to agentic AI systems that can plan over the long term and utilize tools in context. Essentially, they not only advise on "what" needs to be done, but also strategize on "how" to achieve it by selecting and coordinating the right tools. Consider them the cognitive engines driving the upcoming wave of automation—able to analyze on-chain data, perform calculations, create visual representations, and run simulations in a flexible, goal-driven manner.