Model comparison
GPT-5.1-Codex vs GPT-5.6 Luna
GPT-5.6 Luna is the stronger model overall, scoring 54.7 to 38.2 on the Noometry Index.
Last verified . 3 shared benchmarks.
Summary
- They share 3 benchmarks with published results for both. GPT-5.1-Codex scores higher in 1 category and GPT-5.6 Luna in 2 categories; 3 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.6 Luna leads 77.5 to 29.3.
- The biggest single-benchmark swing is ProofBench: 9% for GPT-5.1-Codex and 60% for GPT-5.6 Luna.
- GPT-5.6 Luna is cheaper at $0.20 / $1.20 per million input/output tokens, against $1.25 / $10 for GPT-5.1-Codex.
- GPT-5.6 Luna accepts more context: 1.05M tokens versus 400K.
Side by side
| GPT-5.1-Codex | GPT-5.6 Luna | |
|---|---|---|
| Provider | OpenAI | OpenAI |
| Noometry Index | 38.2 | 54.7 |
| Released | 2025-11-12 | 2026-07-09 |
| Weights | Proprietary | Proprietary |
| Context window | 400K | 1.05M |
| Max output | 128K | 128K |
| Input $ / M tokens | $1.25 | $0.20 |
| Output $ / M tokens | $10 | $1.20 |
| Results tracked | 6 | 51 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5.6 Luna leads
GPT-5.1-Codex: 41.7 (#107), GPT-5.6 Luna: 53.9 (#25)
| Benchmark | GPT-5.1-Codex | GPT-5.6 Luna |
|---|---|---|
| LMArena WebDev | 1337 | 1517 |
| ALE-Bench | 1,245 | 1,667 |
| DeepSWE | — | 67.2% |
| FrontierCode | — | 39.8% |
| SWE-bench Verified (bash only) | 66% | — |
| CursorBench | — | 35.9% |
| SciCode | — | 53.6% |
| WeirdML | — | 60.9% |
| LMArena Coding | — | 1466 |
Agentic & Tool Use GPT-5.1-Codex leads
GPT-5.1-Codex: 37.8 (#32), GPT-5.6 Luna: 35.3 (#37)
| Benchmark | GPT-5.1-Codex | GPT-5.6 Luna |
|---|---|---|
| Terminal-Bench | 60.4% | — |
| BALROG | — | 45.6% |
| GDP.pdf | — | 22.7% |
| METR Time Horizons | 70.8% | — |
| Vending-Bench 2 | — | 4,095 |
Reasoning Not comparable
GPT-5.1-Codex: —, GPT-5.6 Luna: 47.5 (#38)
| Benchmark | GPT-5.1-Codex | GPT-5.6 Luna |
|---|---|---|
| ARC-AGI-2 | — | 59.5% |
| SimpleBench | — | 46.8% |
| Kagi LLM Benchmark | — | 49.1% |
| NYT Connections (extended) | — | 69.4% |
| ARC-AGI-1 | — | 88% |
| CritPt | — | 20.6% |
| Chess Puzzles | — | 40% |
| LMArena Hard Prompts | — | 1452 |
| Mystery Game Puzzles | — | 21% |
| DTBench | — | 89.1% |
| LMCA | — | 48.5% |
| Surface Evolver Bench | — | 61.9% |
| Epoch Capabilities Index | — | 156.32 |
Math GPT-5.6 Luna leads
GPT-5.1-Codex: 29.3 (#218), GPT-5.6 Luna: 77.5 (#12)
| Benchmark | GPT-5.1-Codex | GPT-5.6 Luna |
|---|---|---|
| ProofBench | 9% | 60% |
| FrontierMath (Tiers 1-3) | — | 82.1% |
| FrontierMath Tier 4 | — | 61% |
| OTIS Mock AIME 2024-2025 | — | 98.3% |
| LMArena Math | — | 1455 |
Knowledge Not comparable
GPT-5.1-Codex: —, GPT-5.6 Luna: 58.6 (#31)
| Benchmark | GPT-5.1-Codex | GPT-5.6 Luna |
|---|---|---|
| GPQA Diamond | — | 91.6% |
| SimpleQA Verified | — | 41% |
| LMArena Expert | — | 1479 |
Multimodal Not comparable
GPT-5.1-Codex: —, GPT-5.6 Luna: 43.0 (#22)
| Benchmark | GPT-5.1-Codex | GPT-5.6 Luna |
|---|---|---|
| LMArena Vision | — | 1259 |
| Blueprint-Bench 2 | — | 22.6% |
| Furniture Assembly | — | 42.5% |
| LMArena Document | — | 1457 |
Multilingual Not comparable
GPT-5.1-Codex: —, GPT-5.6 Luna: 53.0 (#73)
| Benchmark | GPT-5.1-Codex | GPT-5.6 Luna |
|---|---|---|
| LMArena Non-English | — | 1418 |
| LMArena Chinese | — | 1469 |
| LMArena French | — | 1457 |
| LMArena German | — | 1459 |
| LMArena Japanese | — | 1416 |
| LMArena Korean | — | 1414 |
| LMArena Russian | — | 1428 |
| LMArena Spanish | — | 1445 |
Instruction Following Not comparable
GPT-5.1-Codex: —, GPT-5.6 Luna: 76.0 (#48)
| Benchmark | GPT-5.1-Codex | GPT-5.6 Luna |
|---|---|---|
| LMArena Instruction Following | — | 1440 |
Long Context Not comparable
GPT-5.1-Codex: —, GPT-5.6 Luna: 43.9 (#78)
| Benchmark | GPT-5.1-Codex | GPT-5.6 Luna |
|---|---|---|
| LMArena Longer Query | — | 1437 |
Writing & Preference Not comparable
GPT-5.1-Codex: —, GPT-5.6 Luna: 68.3 (#27)
| Benchmark | GPT-5.1-Codex | GPT-5.6 Luna |
|---|---|---|
| LMArena Text | — | 1432 |
| LMArena Creative Writing | — | 1397 |
| EQ-Bench Creative Writing | — | 1829 |
| EQ-Bench 4 | — | 1156 |
| LMArena Multi-Turn | — | 1433 |
Frequently asked questions
Is GPT-5.1-Codex better than GPT-5.6 Luna?
GPT-5.6 Luna is the stronger model overall, scoring 54.7 to 38.2 on the Noometry Index.
Which is cheaper, GPT-5.1-Codex or GPT-5.6 Luna?
GPT-5.6 Luna is cheaper. It lists at $0.20 per million input tokens and $1.20 per million output tokens; GPT-5.1-Codex lists at $1.25 and $10.
Is GPT-5.1-Codex or GPT-5.6 Luna better for coding?
GPT-5.6 Luna scores higher on coding benchmarks: 53.9 versus 41.7 in the Noometry coding category.
Which has the bigger context window?
GPT-5.6 Luna does, with 1.05M tokens against 400K.
How many benchmarks do GPT-5.1-Codex and GPT-5.6 Luna share?
3 benchmarks have published results for both models. GPT-5.1-Codex has 6 scored results on Noometry and GPT-5.6 Luna has 51.