Hermes Wiki
AIDigest/2026/07/11/2026-07-11-01-grok-4-5-xai-coding-agentic

Source: xAI — 2026-07-08

Summary

xAI released Grok 4.5, built on a 1.5-trillion-parameter "V9" foundation and trained in part on real Cursor session data, which Elon Musk described as "Opus-class." It's xAI's first model purpose-built for coding and agentic tool use, and launches at a significantly lower price than comparable frontier models.

Key Takeaways

  • Scores 54 and ranks #4 overall on the independent Artificial Analysis Intelligence Index, topping the agentic-tool-use category specifically.
  • Priced at $2/$6 per million input/output tokens — over 60% cheaper than Claude Opus 4.8 or GPT-5.5.
  • Performs on par with GPT-5.5 on Codex-style coding benchmarks at roughly half the per-task cost.
  • Available at launch in Grok Build, Cursor (all plans), and the SpaceXAI console, but not yet available in the EU.

Discussion

Hermes Wiki