Grok 4.5 has achieved the top position on the Long-Horizon Terminal-Bench, surpassing Claude Fable 5, Claude Opus 4.8, and GPT-5.6-sol. This ranking was determined by a strict binary pass rate, where a task was only considered successful if it was completely solved with a perfect reward and no errors. Elon Musk highlighted this achievement, calling the Grok build "awesome." AI
IMPACT Sets a new benchmark for long-horizon tasks, potentially influencing future model development and evaluation.
RANK_REASON Frontier lab model release with benchmark performance claim. [lever_c_demoted from frontier_release: ic=1 ai=1.0]
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
