Model comparison
GPT-5.6 Luna vs Mercury 2.5
GPT-5.6 Luna is the stronger model overall, scoring 54.6 to 33.5 on the Noometry Index. Mercury 2.5 costs 6.7× less per token, which makes it the better buy when GPT-5.6 Luna's lead doesn't matter for your workload.
Last verified . 4 shared benchmarks.
Summary
- They share 4 benchmarks with published results for both. GPT-5.6 Luna scores higher in 3 categories and Mercury 2.5 in 0 categories; 3 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.6 Luna leads 77.7 to 23.3.
- The biggest single-benchmark swing is ProofBench: 60% for GPT-5.6 Luna and 3% for Mercury 2.5.
- Mercury 2.5 is cheaper at $0.04 / $0.15 per million input/output tokens, against $0.20 / $1.20 for GPT-5.6 Luna.
- GPT-5.6 Luna accepts more context: 1.05M tokens versus 260K.
Side by side
| GPT-5.6 Luna | Mercury 2.5 | |
|---|---|---|
| Provider | OpenAI | Inception |
| Noometry Index | 54.6 | 33.5 |
| Released | 2026-07-09 | 2026-09-08 |
| Weights | Proprietary | Proprietary |
| Context window | 1.05M | 260K |
| Max output | 128K | 66K |
| Input $ / M tokens | $0.20 | $0.04 |
| Output $ / M tokens | $1.20 | $0.15 |
| Results tracked | 52 | 4 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5.6 Luna leads
GPT-5.6 Luna: 54.5 (#28), Mercury 2.5: 39.5 (#156)
| Benchmark | GPT-5.6 Luna | Mercury 2.5 |
|---|---|---|
| SciCode | 53.6% | 38.5% |
| ALE-Bench | 1,667 | 301.65 |
| DeepSWE | 67.2% | — |
| FrontierCode | 39.8% | — |
| CursorBench | 35.9% | — |
| LMArena WebDev | 1519 | — |
| WeirdML | 60.9% | — |
| LMArena Coding | 1466 | — |
Agentic & Tool Use Not comparable
GPT-5.6 Luna: 34.4 (#45), Mercury 2.5: —
| Benchmark | GPT-5.6 Luna | Mercury 2.5 |
|---|---|---|
| APEX-Agents | 43% | — |
| BALROG | 45.6% | — |
| GDP.pdf | 22.7% | — |
| Vending-Bench 2 | 4,095 | — |
Reasoning GPT-5.6 Luna leads
GPT-5.6 Luna: 47.6 (#43), Mercury 2.5: 22.4 (#193)
| Benchmark | GPT-5.6 Luna | Mercury 2.5 |
|---|---|---|
| CritPt | 20.6% | 0% |
| ARC-AGI-2 | 59.5% | — |
| SimpleBench | 46.8% | — |
| Kagi LLM Benchmark | 49.1% | — |
| NYT Connections (extended) | 69.4% | — |
| ARC-AGI-1 | 88% | — |
| Chess Puzzles | 40% | — |
| LMArena Hard Prompts | 1451 | — |
| Mystery Game Puzzles | 21% | — |
| DTBench | 89.1% | — |
| LMCA | 48.5% | — |
| Surface Evolver Bench | 61.9% | — |
| Epoch Capabilities Index | 156.39 | — |
Math GPT-5.6 Luna leads
GPT-5.6 Luna: 77.7 (#14), Mercury 2.5: 23.3 (#272)
| Benchmark | GPT-5.6 Luna | Mercury 2.5 |
|---|---|---|
| ProofBench | 60% | 3% |
| FrontierMath (Tiers 1-3) | 82.1% | — |
| FrontierMath Tier 4 | 61% | — |
| OTIS Mock AIME 2024-2025 | 98.3% | — |
| LMArena Math | 1458 | — |
Knowledge Not comparable
GPT-5.6 Luna: 58.5 (#34), Mercury 2.5: —
| Benchmark | GPT-5.6 Luna | Mercury 2.5 |
|---|---|---|
| GPQA Diamond | 91.6% | — |
| SimpleQA Verified | 41% | — |
| LMArena Expert | 1478 | — |
Multimodal Not comparable
GPT-5.6 Luna: 42.7 (#28), Mercury 2.5: —
| Benchmark | GPT-5.6 Luna | Mercury 2.5 |
|---|---|---|
| LMArena Vision | 1258 | — |
| Blueprint-Bench 2 | 22.6% | — |
| Furniture Assembly | 42.5% | — |
| LMArena Document | 1457 | — |
Multilingual Not comparable
GPT-5.6 Luna: 52.8 (#78), Mercury 2.5: —
| Benchmark | GPT-5.6 Luna | Mercury 2.5 |
|---|---|---|
| LMArena Non-English | 1417 | — |
| LMArena Chinese | 1470 | — |
| LMArena French | 1456 | — |
| LMArena German | 1454 | — |
| LMArena Japanese | 1411 | — |
| LMArena Korean | 1415 | — |
| LMArena Russian | 1428 | — |
| LMArena Spanish | 1448 | — |
Instruction Following Not comparable
GPT-5.6 Luna: 75.6 (#57), Mercury 2.5: —
| Benchmark | GPT-5.6 Luna | Mercury 2.5 |
|---|---|---|
| LMArena Instruction Following | 1437 | — |
Long Context Not comparable
GPT-5.6 Luna: 43.9 (#82), Mercury 2.5: —
| Benchmark | GPT-5.6 Luna | Mercury 2.5 |
|---|---|---|
| LMArena Longer Query | 1436 | — |
Writing & Preference Not comparable
GPT-5.6 Luna: 68.0 (#29), Mercury 2.5: —
| Benchmark | GPT-5.6 Luna | Mercury 2.5 |
|---|---|---|
| LMArena Text | 1431 | — |
| LMArena Creative Writing | 1396 | — |
| EQ-Bench Creative Writing | 1829 | — |
| EQ-Bench 4 | 1156 | — |
| LMArena Multi-Turn | 1434 | — |
Frequently asked questions
Is GPT-5.6 Luna better than Mercury 2.5?
GPT-5.6 Luna is the stronger model overall, scoring 54.6 to 33.5 on the Noometry Index. Mercury 2.5 costs 6.7× less per token, which makes it the better buy when GPT-5.6 Luna's lead doesn't matter for your workload.
Which is cheaper, GPT-5.6 Luna or Mercury 2.5?
Mercury 2.5 is cheaper. It lists at $0.04 per million input tokens and $0.15 per million output tokens; GPT-5.6 Luna lists at $0.20 and $1.20.
Is GPT-5.6 Luna or Mercury 2.5 better for coding?
GPT-5.6 Luna scores higher on coding benchmarks: 54.5 versus 39.5 in the Noometry coding category.
Which has the bigger context window?
GPT-5.6 Luna does, with 1.05M tokens against 260K.
How many benchmarks do GPT-5.6 Luna and Mercury 2.5 share?
4 benchmarks have published results for both models. GPT-5.6 Luna has 52 scored results on Noometry and Mercury 2.5 has 4.