I tested the 3 Claude Code alternatives the internet keeps talking about, and only one let me down

GitHub has announced that Copilot Fusion is now generally available, a system designed to handle model selection and orchestration for coding tasks. Debuted at GitHub Universe in San Francisco, the system allows developers to select an effort level rather than a specific model. These levels include Fusion Standard for everyday coding, Fusion High for harder tasks, and Fusion Max for the most demanding work.

According to The New Stack, Fusion Max delivers benchmark performance comparable to Claude Code with Opus 5.5 Max, but at approximately half the average API cost and in roughly 40% less time per task. The system utilizes three different workflows: Single, where one model solves a task; Cascade, which uses an efficient model that can escalate to a stronger model; and Critique, where one model drafts and another reviews.

While the Copilot plans page lists models from providers such as Anthropic, OpenAI, Google, xAI, and Microsoft, Dev.to reports that GitHub does not publish which specific models Fusion selects. XDA Developers notes that Codex, Antigravity, and OpenCode are frequently discussed on platforms like Reddit and YouTube as potential alternatives to Claude Code.

Brief written by urgent.news from XDA Developers, Dev.to, The New Stack — 3 reports on this story. Machine-written — may contain errors; check the original before relying on it.