GPT-5.6 Luna vs Muse Spark 1.1
GPT-5.6 Luna leads Muse Spark 1.1 by 3.1 points on the Shogo Agent Index, but their ranges overlap, so treat it as close. GPT-5.6 Luna wins 5 and Muse Spark 1.1 wins 6 of 11 shared benchmarks. GPT-5.6 Luna is cheaper for agent workloads.
Add more models to this comparison →Pillars
| Pillar | GPT-5.6 Luna | Muse Spark 1.1 |
|---|---|---|
| Agentic coding | 59 | 53 |
| Long-horizon autonomy | 61 | 64 |
| Knowledge-work agents | 39 | 34 |
| Computer use | 40 | 36 |
| Security & context learning | 51 | 48 |
Shared benchmarks
| Benchmark | GPT-5.6 Luna | Muse Spark 1.1 |
|---|---|---|
| APEX-Agentsindex | 43.0% | 31.8% |
| DeepSWEindex | 67.2% | 53.3% |
| Vending-Bench 2index | $4,095 | $6,520 |
| Chess Puzzles | 43.0% | 34.0% |
| CritPt | 20.6% | 15.1% |
| DTBench | 89.1% | 94.4% |
| LMCA | 48.5% | 49.9% |
| ProofBench | 60.0% | 39.0% |
| SciCode | 53.6% | 58.8% |
| SimpleQA Verified | 41.0% | 57.8% |
| WebDev Arena | 1519 | 1542 |
Specs and price
| GPT-5.6 Luna | Muse Spark 1.1 | |
|---|---|---|
| Agent Index | 51.4 (#34) | 48.3 (#40) |
| Agent $/M | $0.10 | $0.50 |
| Input / output $/M | $0.20 / $1.20 | $1.25 / $4.25 |
| Context | 1.1M | 1M |
| Released | Jul 9, 2026 | Apr 8, 2026 |
| Weights | Closed | Closed |
| API hosts | 5 | 1 |
Don’t pick one. Route.
Shogo can run GPT-5.6 Luna and Muse Spark 1.1 side by side and send each step of an agent to the one that’s cheaper for it.
Try both in Shogo