Claude Code GA + Antigravity 2.0 unified harness — the H2 2026 coding-agent landscape stratifies into CLI-first, IDE-first, and unified-surface positions

Claude Code today, CLI-first $20/month bundled into Claude Pro. Antigravity 2.0 with unified harness for both IDE and CLI surfaces. The H2 2026 coding-agent landscape stratifies into three structural positions: CLI-first (Claude Code, OpenCode terminal), IDE-first (Cursor, Copilot, Junie), unified-surface (Antigravity). Procurement decisions match workflow preference.

Claude Code's CLI-first GA and Antigravity 2.0's unified-harness architecture together represent the H2 2026 coding-agent landscape stratification. Pre-2026 most coding agents targeted either IDE-first (Cursor pattern) or CLI-first (OpenCode terminal pattern). The H2 2026 landscape adds unified-surface (Antigravity covering both) as a third structural position.

The CLI-first procurement segment

Infrastructure-and-devops engineers who operate primarily from terminal increasingly choose CLI-first coding agents (Claude Code, OpenCode terminal mode) over IDE-integrated alternatives. The workflow fit matters substantially — IDE-based agents that don't integrate with terminal workflows underperform in CLI-heavy work patterns. Claude Code's $20/month bundled-Claude-Pro pricing makes the CLI-first option substantially cheaper than standalone competitors like Cognition Devin.

The IDE-first procurement segment

Application-development engineers who operate primarily from IDEs (VS Code, JetBrains, etc.) increasingly choose IDE-first coding agents. Cursor 3, GitHub Copilot Desktop, JetBrains Junie all target this segment with different specializations. The competitive intensity in this segment is high but the workflow-fit advantage IDE-first agents have over CLI-first in this segment is structural.

The unified-surface bet

Google's Antigravity 2.0 unified-harness architecture attempts to cover both segments through one platform. The H2 2026 procurement question is whether the unified approach wins both segments or falls into 'jack-of-all-trades-master-of-none' territory. Each segment has specialized vendors that fit the segment-specific workflow more precisely than the unified approach. The H2 2026 procurement-evaluation outcomes will surface which structural position actually wins.

Stackademic — 5 AI Coding Agents That Actually Ship Production Code in 2026 → · MorphLLM — Best AI Coding Agents (June 2026): Scored Leaderboard →