GPT-5.6 Luna vs GPT-6.1 Terra
GPT-6.1 Terra leads GPT-5.6 Luna by 17.6 points on the Shogo Agent Index. GPT-5.6 Luna wins 0 and GPT-6.1 Terra wins 12 of 12 shared benchmarks. GPT-5.6 Luna is cheaper for agent workloads.
Add more models to this comparison →Pillars
| Pillar | GPT-5.6 Luna | GPT-6.1 Terra |
|---|---|---|
| Agentic coding | 59 | 79 |
| Long-horizon autonomy | 61 | 74 |
| Knowledge-work agents | 39 | 56 |
| Computer use | 40 | 63 |
| Security & context learning | 51 | 62 |
Shared benchmarks
| Benchmark | GPT-5.6 Luna | GPT-6.1 Terra |
|---|---|---|
| APEX-Agentsindex | 43.0% | 68.3% |
| Vending-Bench 2index | $4,095 | $9,107 |
| CritPt | 20.6% | 28.0% |
| FrontierMath Tier 4 | 61.0% | 92.7% |
| FrontierMath Tiers 1-3 | 82.1% | 90.9% |
| GPQA Diamond | 91.6% | 94.3% |
| MirrorCode | 21.1% | 76.7% |
| OTIS Mock AIME | 98.3% | 100.0% |
| SciCode | 53.6% | 55.1% |
| SimpleBench | 46.8% | 72.2% |
| SimpleQA Verified | 41.0% | 57.3% |
| WeirdML v3 | 5.2% | 28.1% |
Specs and price
| GPT-5.6 Luna | GPT-6.1 Terra | |
|---|---|---|
| Agent Index | 51.4 (#34) | 69 (#8) |
| Agent $/M | $0.10 | $0.42 |
| Input / output $/M | $0.20 / $1.20 | $1.00 / $5.00 |
| Context | 1.1M | 1.1M |
| Released | Jul 9, 2026 | Oct 26, 2026 |
| Weights | Closed | Closed |
| API hosts | 5 | 4 |
Don’t pick one. Route.
Shogo can run GPT-5.6 Luna and GPT-6.1 Terra side by side and send each step of an agent to the one that’s cheaper for it.
Try both in Shogo