Claude Opus 5 is a large language model in Anthropic's Claude family. Recent coverage focuses on benchmark and usage results: on the ThinkingBox-Bench of 507 business workflows, it solved 79.09% of tasks at least once and repeated success across all 20 runs 47.53% of the time, ahead of Kimi-K3, and it was also tested through Claude Code on OpenSCAD versus CadQuery CAD tasks. The later Claude Opus 5.5 is described as generating output tokens over 30% faster, with em dashes falling from 15.2 to 0.8 per 1,000 words.
ThinkingBox benchmark: one-shot agent success rarely repeatsMicrosoft authors released ThinkingBox-Bench: 507 business workflows across 5 domains, each run 20 times and graded on terminal backend state. Kimi-K3 solved 93.89% of tasks at least once but only 13.41% all 20 times, while Claude Opus 5 discovered fewer (79.09%) yet repeated far more (47.53%), showing discovery and repeatability rank models almost oppositely.
Anthropic Releases Claude Opus 5.5 Prompting GuideAnthropic released a prompting guide for Claude Opus 5.5, which generates output tokens over 30% faster than Opus 5 and tends to finish tasks with fewer tokens. Existing Opus 5 prompts should work without changes; the guide covers agentic coding, multi-app workflows and safeguard refusals.
Claude Opus 5.5 uses 95% fewer em dashes but longer answersArena analysis shows Anthropic's Claude Opus 5.5 reduced telltale AI writing patterns versus Opus 5: em dashes fell from 15.2 to 0.8 per 1,000 words (about 95% fewer), semicolons from 6.10 to 1.64, and average sentence length from 12.14 to 10.03 words, though answers grew from 453 to 481 words.
Benchmark: CadQuery vs OpenSCAD for agentic CAD workModelRift ran a controlled comparison of OpenSCAD versus CadQuery (OpenCascade B-rep) for fully agent-driven CAD tasks. Six agents solved three tasks using Claude Opus 5 through Claude Code; all resulting STLs were printable, but the tools differ in how they fail. Tests used CadQuery 2.8.0 on Python 3.14 and OpenSCAD 2026.06.12 on an M-series Mac; output sizes were ~13.1 KB for CadQuery and ~12.4 KB for OpenSCAD.
The A.I. Daily