BREAKING
Grok 4.5 Draws Level With ChatGPT
GDPval+ Professional Tasks
Grok 4.529
GPT-5.522
Opus 4.821
Coding Benchmarks Split the Lead
SWE-Atlas Grok84
SWE-Atlas GPT-5.684
SWE-bench Grok69.1
SWE-bench GPT74.9
Two Complementary Systems
Grok 4.5$30/mo
Real-time X and news
Faster output
2M-token context
ChatGPT$20/mo
Polished writing
Enterprise reliability
Mature ecosystem
Leader Now Depends on the Task
The Gap Has Narrowed
AI NEWS BLITZ
xAI's Grok 4.5 now matches or beats OpenAI's flagship models on key benchmarks.