Pine has developed a dedicated AI computer: it ranks first in benchmark tests, but lags behind Codex and Claude Code in real-world tasks.

Beating AI Express News reports that Pine AI has launched Pine Computer, a cloud computer designed specifically for AI Agents. It integrates cloud virtual machines, AI models, and operational tools. Developers submit tasks via API, enabling the AI to operate browsers, software, and files in the cloud without requiring users to use their own devices. Pine’s core improvement is enabling the computer to proactively inform the AI of changes, addressing a key limitation of traditional vision-based Agents, which typically need repeated screenshots to assess click outcomes. Pine instead lets browsers and operating systems directly feed back changes to the AI model executing the task, cutting down on repeated screenshots and page checks. Pine claims its new solution completes tasks 2 to 5 times faster than traditional approaches in internal tests. In the SaaS-Bench v1.1 test, Pine’s checkpoint score hit 78.3%, outperforming Claude Code’s 74.3% and Codex’s 71.1%. This score measures how many check requirements are fulfilled during task execution, not overall task success rate. However, X Community notes point out Pine’s full task completion rate is only 27.4%, lower than Claude Code’s 31.1% and Codex’s 29.2%. Pine’s model costs are also far lower, averaging around $1.02 per task, versus $20.50 for Codex and $26.50 for Claude Code. That said, these fees exclude cloud server costs, and each solution uses different models and execution budgets. Currently, Pine Computer remains in an invitation-only closed beta phase.