xAI Grok 4.5 Launch
2 videos · 2 channels · score 2.7k
xAI unveiled Grok 4.5, a coding‑focused LLM trained on data from the acquired Cursor tool, and creators like Wes Roth and Matthew Berman hail it as on par with GPT‑5.5 Extra High and five points ahead of Opus 4.8 in agentic coding benchmarks—demonstrated by a 3D ship‑sailing game with realistic physics—sparking buzz over its claim to outpace the current AI coding elite.
The coverage — 2 videos
Grok 4.5 just COOKED Claude and OpenAI
xAI launched Grok 4.5 today, trained with Cursor acquisition data, claiming it matches GPT-5.5 Extra High and Opus 4.8 Max on coding benchmarks and demonstrated by creating a 3D ship sailing game with realistic physics and audio.
Grok just broke the trend
Grok 4.5 has surpassed Opus 4.8 in benchmark scores, with a significant five-point lead in agentic coding benchmarks, as revealed in the latest benchmark scores release