Cursor user finds grok-4.7 performs much better with gpt-5.6-sol as orchestrator
letandrewcook · x · 2026-11-03
A Cursor user shared his usage dashboard: 1.8bn tokens burned this cycle (98.7% on grok-4.7-high) plus 166.4m tokens of gpt-5.6-sol-medium. His finding: grok-4.7 can be really good if you use gpt-5.6-sol as your orchestrator or "chief of staff."
A non-obvious multi-model combo inside a coding tool — one model for planning/coordination, the other for execution.
More from coding & agent
- Berkeley's GEPA-Curator filters noisy failure signals to boost agent prompt optimization — Berkeley · 2026-11-03
- Default AI agents to serverless, then let production data decide what moves — rseroter · 2026-11-03
- Skill-centric training beats task-centric for small web agents across 4 benchmarks — anmarasovic · 2026-11-03
- Your agentic coding trajectories are a treasure, argues new blog post — Bartaseth · 2026-11-03
- Dev argues agents make experimentation so cheap you should be 10x more scientific — zack_overflow · 2026-11-03
- Anthropic's message threads cut TTFT 35% on 1M-token agent sessions — ClaudeDevs · 2026-11-03