Grok 4.7 xHigh hits 46.3% on CursorBench 4.0, matching Opus 5 Max at half the cost
XFreeze · x · 2026-09-22
Grok 4.7 xHigh scores 46.3% on CursorBench 4.0, essentially matching Claude Opus 5 Max (46.6%) while costing roughly half per task:
- Opus 5 Max → 46.6% · $11.95/task · 85k tokens · 106 steps
- Grok 4.7 xHigh → 46.3% · $6.01/task · 70k tokens · 88 steps
- GPT-5.6 Sol Max → 41.7% · $8.23/task · 43k tokens · 99 steps
Grok 4.7 also keeps the same $2/$6 pricing as Grok 4.6, making it a strong value pick for agentic coding workloads.
Related event: Grok 4.7 debuts near top of benchmarks at lower cost(25 posts)→
More from coding & agent
- Cloudflare launches Worker Previews: isolated production-like envs for every Git branch — iannuttall · 2026-09-22
- The 'chief of staff' pattern: how agentic coding scales past context windows — arpit_bhayani · 2026-09-22
- FlexViz: open-source agent-native data viz where agents read your dashboard as structured state, 1B points in 251ms — Adorable-Giraffe5754 · 2026-09-22
- Jay Alammar and Team to Host Free AI Agents Masterclass on Sept 29 — JayAlammar · 2026-09-22
- Hands-on: testing Grok 4.7 coding in Cursor across 4 real projects with cost breakdown — Arindam_1729 · 2026-09-22
- Agent task complete isn't case closed: who owns the handoff after AI finishes? — IrfanZahoor_950 · 2026-09-22