Fetching from the wire…
01
02
03
04
05
06
07
08
Public story · 2026-02-12 · source-backed
Opus 4.6 (Thinking) leads SWE-bench Verified at 79.2%, but SWE-Bench Pro (~23%) and SWE-EVO (19-21%) reveal significant gaps. Current coding agents excel at well-defined single-repo issues but struggle with cross-language, cross-repo, evolutionary tasks — the reality of production engineering.
Each link below shares sources, entities, or timing with this story.
Claude Code uses Opus / Shared entities / Shared topic / What happened next
Linked by a graph relationship (Claude Code uses Opus); both cover Bench Pro, Opus, SWE, Thinking; overlapping topics (agent, benchmark).
Cursor uses Opus / Shared entities / Shared topic / What happened next / Tension
Linked by a graph relationship (Cursor uses Opus); both cover Bench Pro, Opus, SWE; overlapping topics (benchmark, opus).
Claude Code uses Opus / Shared entities / Shared topic / What happened next / Tension
Linked by a graph relationship (Claude Code uses Opus); both cover Opus, SWE, Verified; overlapping topics (coding, swe-bench).
Opus built by Anthropic / Shared entities / Shared topic / What happened next / Tension
Linked by a graph relationship (Opus built by Anthropic); both cover Bench Pro, Opus, SWE; overlapping topics (agent, benchmark, coding, swe-bench).
Claude Code uses Opus / Shared entities / Shared topic / What happened next / Tension
Linked by a graph relationship (Claude Code uses Opus); both cover Opus, SWE, Verified; overlapping topics (benchmark, coding, opus).
Opus built by Anthropic / Shared entities / Shared topic / What happened next
Linked by a graph relationship (Opus built by Anthropic); both cover Bench Pro, Opus, SWE; overlapping topics (agent, benchmark, coding, engineering, opus).
Cursor uses Opus / Shared entities / Shared topic / What happened next
Linked by a graph relationship (Cursor uses Opus); both cover Opus, SWE; overlapping topics (agent, coding, opus).
Linked by a graph relationship (Cursor uses Opus); both cover SWE, Verified; overlapping topics (opus, swe-bench).