Grok 4.5 Launched with Benchmark Scores

shensi · x · 2026-07-09

Merge Gateway launches Grok 4.5, offering it as a new model for coding and agentic tasks.

The original post also notes that the model scored 83.3% on Terminal Bench 2.1, highlighting model availability changes and real-world performance.

Related event: xAI Launches Grok 4.5: Coding and Agent Focus to Rival Opus(55 posts)→

Original post →

More from coding & agent

coding & agent channel →