Expired75% confidenceEventTime unknown
Developer Henry-Jessie used GPT-5.2 together with the Codex CLI tool to achieve a score surpassing Claude Opus 4.5 in Anthropic's Performance Take-Home challenge.
1
Sources
75%
Confidence
Medium-term (~90 days)
Relevance
6/1/2026
First Seen
Valid until: 8/30/2026(expired)
Sources
Related Entities
Related Claims
UnverifiedThe Skill is based on Codex and uses GPT-5.x-level models to complete the publishing form-filling for multi-platform video content.71% similarUnverifiedA reviewer named Evand retested Codex for two weeks using GPT 5.5 and concluded that Codex is already in the same tier as Claude Code on many tasks, and is faster and more accurate in certain scenarios.71% similarUnverifiedOpus 4.5 in High mode responds significantly faster than GPT-5.2 Codex and supports deep integration with the IDE environment via a VS Code plugin.71% similarUnverifiedCodex CLI currently supports the GPT-5 model and offers 4 different levels of thinking depth.70% similarUnverifiedGPT-5 Codex's web version delivered the best results across the four tests, with soft-body physics deformation effects close to the original game.70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/30281API
curl https://kongchang.com/api/v1/knowledge/claims/30281MCP
get_claim(id=30281)