07:34
2026-08-14
runtimewire.com
artificial-intelligence
Grok 4.6 tops Newsroom Reliability v0.2 benchmark at 0.79
SpaceXAI's Grok 4.6 topped the Newsroom Reliability v0.2 benchmark with a mean score of 0.79 across 50 tasks at an estimated $0.0056 per task, according to RuntimeWire's leaderboard. OpenAI's GPT-5.6 โฆ