en.Wedoany.com Reported - GitHub is extending Copilot from the integrated development environment (IDE) to the broader software development lifecycle, launching a new desktop application and a collaborative workbench called canvas, aiming to transform the AI-assisted coding tool into a "control center" for agent-native software development.

The company announced at Microsoft's annual Build conference that the desktop application is designed to provide developers with a dedicated environment for collaborating with AI agents throughout the entire software development lifecycle. The application includes a collaborative workspace called canvas, where developers can brainstorm, refine requirements, plan, and iterate on projects with AI. It also features new Agent Merge and code review capabilities, enabling developers to automate Copilot to merge tasks from different agents to achieve specific goals, or conduct autonomous code reviews based on established criteria. Phil Fersht, CEO of HFS Research, believes these new features can reduce context switching, improve engineering efficiency, and accelerate delivery cycles.

However, despite these new features, recent developer discussions have largely focused on another topic: GitHub Copilot will shift to a usage-based pricing model this week (announced by the company in April). This change has drawn criticism on the GitHub Community Forum, with users accusing the company of "bait and switch," and some requesting refunds or announcing plans to cancel subscriptions.

For analysts, this pricing adjustment is necessary and reasonable from GitHub's perspective. Advait Patel, Senior Reliability Engineer at Broadcom, stated: "The justification for the pricing adjustment lies in GitHub's future direction, not the current product state. Running multiple agents in parallel, combined with sandboxes, canvas reviews, and Agent Merge through CI loops, is closer to cloud computing than an IDE plugin, and therefore cannot be priced on a fixed seat basis. Structurally, usage-based pricing is the right choice." Fersht noted that developers and CIOs need to focus on Copilot's transformation from a coding assistant to a platform for orchestrating software development agents and workflows, shifting metrics from "lines of code generated" to broader operational outcomes such as release speed, code quality, defect reduction, and engineering efficiency.

GitHub is not the only vibe-coding tool company rethinking its pricing strategy. Over the past year, platforms like Claude Code, Replit, Cursor, and Kiro have adjusted their pricing structures multiple times to address rising infrastructure costs, limited GPU availability, and the overhead of serving increasingly complex AI models and agents. Amit Chandak, Chief Analytics Officer at IT consulting firm Kanerika, believes developers and CIOs should focus less on GitHub's pricing mechanism and more on whether these tools deliver measurable business value. He said: "The new features announced by GitHub can serve as productivity multipliers, but they could also be features that increase consumption without delivering corresponding business value. If productivity baselines are not established before adoption, enterprises may incur higher costs without clearly tracing the value delivered." Fersht believes that as pricing models change, developers and CIOs need to focus on governance, monitoring, and financial controls. "Autonomous agents continuously reason, test, revise, and interact with multiple systems, making their consumption patterns far more unpredictable than traditional SaaS tools."

Patel advises users and decision-makers to maintain more skepticism, especially since the new features are still in technical preview. He added: "Customers are now being asked to pay variable rates for value that has not yet been validated in production. Don't assume new features justify higher spending. Instead, run a 90-day pilot project, measure the number of PRs merged per dollar before and after, and let the data speak. If the ratio improves, the pricing is fair; if not, you are paying for promises, not delivery."

This article is compiled by Wedoany. All AI citations must indicate the source as "Wedoany". If there is any infringement or other issues, please notify us promptly, and we will modify or delete it accordingly. Email: news@wedoany.com