Model comparison
Gemini 2.5 Flash vs GPT-5.6 Luna
GPT-5.6 Luna is the stronger model overall, scoring 54.7 to 39.0 on the Noometry Index.
Last verified . 32 shared benchmarks.
GoogleSummary
- They share 32 benchmarks with published results for both. Gemini 2.5 Flash scores higher in 1 category and GPT-5.6 Luna in 9 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.6 Luna leads 77.5 to 38.1.
- The biggest single-benchmark swing is ARC-AGI-2: 2.5% for Gemini 2.5 Flash and 59.5% for GPT-5.6 Luna.
- GPT-5.6 Luna is cheaper at $0.20 / $1.20 per million input/output tokens, against $0.30 / $2.50 for Gemini 2.5 Flash.
- GPT-5.6 Luna accepts more context: 1.05M tokens versus 1.05M.
Side by side
| Gemini 2.5 Flash | GPT-5.6 Luna | |
|---|---|---|
| Provider | OpenAI | |
| Noometry Index | 39.0 | 54.7 |
| Released | 2025-04-17 | 2026-07-09 |
| Weights | Proprietary | Proprietary |
| Context window | 1.05M | 1.05M |
| Max output | 66K | 128K |
| Input $ / M tokens | $0.30 | $0.20 |
| Output $ / M tokens | $2.50 | $1.20 |
| Results tracked | 54 | 51 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5.6 Luna leads
Gemini 2.5 Flash: 35.5 (#210), GPT-5.6 Luna: 53.9 (#25)
| Benchmark | Gemini 2.5 Flash | GPT-5.6 Luna |
|---|---|---|
| WeirdML | 41.9% | 60.9% |
| LMArena Coding | 1424 | 1466 |
| ALE-Bench | 661.88 | 1,667 |
| DeepSWE | — | 67.2% |
| FrontierCode | — | 39.8% |
| SWE-bench Verified (bash only) | 28.7% | — |
| Aider Polyglot | 55.1% | — |
| CursorBench | — | 35.9% |
| LMArena WebDev | — | 1517 |
| SciCode | — | 53.6% |
Agentic & Tool Use GPT-5.6 Luna leads
Gemini 2.5 Flash: 30.6 (#67), GPT-5.6 Luna: 35.3 (#37)
| Benchmark | Gemini 2.5 Flash | GPT-5.6 Luna |
|---|---|---|
| BALROG | 33.5% | 45.6% |
| Vending-Bench 2 | 548.84 | 4,095 |
| Terminal-Bench | 17.1% | — |
| Berkeley Function Calling Leaderboard | 56.2% | — |
| TheAgentCompany | 41.1% | — |
| GDP.pdf | — | 22.7% |
Reasoning GPT-5.6 Luna leads
Gemini 2.5 Flash: 18.2 (#271), GPT-5.6 Luna: 47.5 (#38)
| Benchmark | Gemini 2.5 Flash | GPT-5.6 Luna |
|---|---|---|
| ARC-AGI-2 | 2.5% | 59.5% |
| SimpleBench | 41.2% | 46.8% |
| Kagi LLM Benchmark | 56.8% | 49.1% |
| ARC-AGI-1 | 33.3% | 88% |
| CritPt | 1.1% | 20.6% |
| LMArena Hard Prompts | 1422 | 1452 |
| DTBench | 76.5% | 89.1% |
| LMCA | 27.5% | 48.5% |
| Epoch Capabilities Index | 143.04 | 156.32 |
| NYT Connections (extended) | — | 69.4% |
| Chess Puzzles | — | 40% |
| EnigmaEval | 2.7% | — |
| Mystery Game Puzzles | — | 21% |
| Surface Evolver Bench | — | 61.9% |
| ForecastBench | 60.6 | — |
Math GPT-5.6 Luna leads
Gemini 2.5 Flash: 38.1 (#87), GPT-5.6 Luna: 77.5 (#12)
| Benchmark | Gemini 2.5 Flash | GPT-5.6 Luna |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 73.1% | 98.3% |
| LMArena Math | 1415 | 1455 |
| FrontierMath (Tiers 1-3) | — | 82.1% |
| FrontierMath Tier 4 | — | 61% |
| ProofBench | — | 60% |
| Omni-MATH | 38.5% | — |
| FrontierMath (Feb 2025 set) | 4.8% | — |
| FrontierMath Tier 4 (v1) | 4.2% | — |
Knowledge GPT-5.6 Luna leads
Gemini 2.5 Flash: 36.1 (#162), GPT-5.6 Luna: 58.6 (#31)
| Benchmark | Gemini 2.5 Flash | GPT-5.6 Luna |
|---|---|---|
| LMArena Expert | 1426 | 1479 |
| GPQA Diamond | — | 91.6% |
| Humanity's Last Exam | 12.1% | — |
| SimpleQA Verified | — | 41% |
| MMLU-Pro | 63.9% | — |
| Confabulations | 16.8% | — |
| Vectara Hallucination Rate | 7.8% | — |
| GPQA (HELM) | 39% | — |
Multimodal GPT-5.6 Luna leads
Gemini 2.5 Flash: 41.2 (#31), GPT-5.6 Luna: 43.0 (#22)
| Benchmark | Gemini 2.5 Flash | GPT-5.6 Luna |
|---|---|---|
| LMArena Vision | 1254 | 1259 |
| GeoBench | 76% | — |
| VPCT | 46.2% | — |
| Blueprint-Bench 2 | — | 22.6% |
| Furniture Assembly | — | 42.5% |
| LMArena Document | — | 1457 |
| SpatialViz-Bench | 36.9% | — |
Multilingual Too close to call
Gemini 2.5 Flash: 52.5 (#83), GPT-5.6 Luna: 53.0 (#73)
| Benchmark | Gemini 2.5 Flash | GPT-5.6 Luna |
|---|---|---|
| LMArena Non-English | 1410 | 1418 |
| LMArena Chinese | 1450 | 1469 |
| LMArena French | 1433 | 1457 |
| LMArena German | 1419 | 1459 |
| LMArena Japanese | 1403 | 1416 |
| LMArena Korean | 1385 | 1414 |
| LMArena Russian | 1415 | 1428 |
| LMArena Spanish | 1421 | 1445 |
Instruction Following Too close to call
Gemini 2.5 Flash: 75.9 (#50), GPT-5.6 Luna: 76.0 (#48)
| Benchmark | Gemini 2.5 Flash | GPT-5.6 Luna |
|---|---|---|
| LMArena Instruction Following | 1405 | 1440 |
| IFEval | 89.8% | — |
Long Context Gemini 2.5 Flash leads
Gemini 2.5 Flash: 47.5 (#16), GPT-5.6 Luna: 43.9 (#78)
| Benchmark | Gemini 2.5 Flash | GPT-5.6 Luna |
|---|---|---|
| LMArena Longer Query | 1419 | 1437 |
| Fiction.LiveBench | 77.8% | — |
Writing & Preference GPT-5.6 Luna leads
Gemini 2.5 Flash: 54.0 (#153), GPT-5.6 Luna: 68.3 (#27)
| Benchmark | Gemini 2.5 Flash | GPT-5.6 Luna |
|---|---|---|
| LMArena Text | 1417 | 1432 |
| LMArena Creative Writing | 1400 | 1397 |
| EQ-Bench Creative Writing | 1137 | 1829 |
| LMArena Multi-Turn | 1408 | 1433 |
| Short-Story Creative Writing | 76.5% | — |
| WildBench | 81.7% | — |
| EQ-Bench 4 | — | 1156 |
Frequently asked questions
Is Gemini 2.5 Flash better than GPT-5.6 Luna?
GPT-5.6 Luna is the stronger model overall, scoring 54.7 to 39.0 on the Noometry Index.
Which is cheaper, Gemini 2.5 Flash 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; Gemini 2.5 Flash lists at $0.30 and $2.50.
Is Gemini 2.5 Flash or GPT-5.6 Luna better for coding?
GPT-5.6 Luna scores higher on coding benchmarks: 53.9 versus 35.5 in the Noometry coding category.
Which has the bigger context window?
GPT-5.6 Luna does, with 1.05M tokens against 1.05M.
How many benchmarks do Gemini 2.5 Flash and GPT-5.6 Luna share?
32 benchmarks have published results for both models. Gemini 2.5 Flash has 54 scored results on Noometry and GPT-5.6 Luna has 51.