GPT-5.6 Luna vs Grok 4.7
Grok 4.7 leads GPT-5.6 Luna by 3.4 points on the Shogo Agent Index, but their ranges overlap, so treat it as close. GPT-5.6 Luna wins 6 and Grok 4.7 wins 14 of 20 shared benchmarks. GPT-5.6 Luna is cheaper for agent workloads.
Add more models to this comparison →Pillars
| Pillar | GPT-5.6 Luna | Grok 4.7 |
|---|---|---|
| Agentic coding | 59 | 58 |
| Long-horizon autonomy | 61 | 69 |
| Knowledge-work agents | 39 | 47 |
| Computer use | 40 | 39 |
| Security & context learning | 51 | 50 |
Shared benchmarks
| Benchmark | GPT-5.6 Luna | Grok 4.7 |
|---|---|---|
| APEX-Agentsindex | 43.0% | 70.0% |
| Vending-Bench 2index | $4,095 | $10,537 |
| CursorBenchindex | 35.9% | 46.3% |
| ARC-AGI | 88.0% | 90.2% |
| ARC-AGI-2 | 59.5% | 61.4% |
| Chess Puzzles | 43.0% | 41.0% |
| CritPt | 20.6% | 18.0% |
| DTBench | 89.1% | 96.0% |
| FrontierCode | 39.8% | 47.6% |
| FrontierMath Tier 4 | 61.0% | 17.1% |
| FrontierMath Tiers 1-3 | 82.1% | 53.0% |
| GPQA Diamond | 91.6% | 92.7% |
| LMCA | 48.5% | 49.4% |
| MirrorCode | 21.1% | 54.4% |
| OTIS Mock AIME | 98.3% | 98.1% |
| ProofBench | 60.0% | 26.0% |
| SciCode | 53.6% | 57.8% |
| SimpleQA Verified | 41.0% | 56.0% |
| WebDev Arena | 1519 | 1638 |
| WeirdML v3 | 5.2% | 8.5% |
Specs and price
| GPT-5.6 Luna | Grok 4.7 | |
|---|---|---|
| Agent Index | 51.4 (#34) | 54.8 (#27) |
| Agent $/M | $0.10 | $0.98 |
| Input / output $/M | $0.20 / $1.20 | $2.00 / $6.00 |
| Context | 1.1M | 500K |
| Released | Jul 9, 2026 | Sep 21, 2026 |
| Weights | Closed | Closed |
| API hosts | 5 | 5 |
Don’t pick one. Route.
Shogo can run GPT-5.6 Luna and Grok 4.7 side by side and send each step of an agent to the one that’s cheaper for it.
Try both in Shogo