Model comparison
Gemini 3.8 Flash vs GPT-6 Luna
Gemini 3.8 Flash is the stronger model overall, scoring 61.8 to 53.3 on the Noometry Index. GPT-6 Luna costs 7.5× less per token, which makes it the better buy when Gemini 3.8 Flash's lead doesn't matter for your workload.
Last verified . 42 shared benchmarks.
GoogleSummary
- They share 42 benchmarks with published results for both. Gemini 3.8 Flash scores higher in 8 categories and GPT-6 Luna in 2 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Gemini 3.8 Flash leads 76.9 to 48.2.
- The biggest single-benchmark swing is Mystery Game Puzzles: 47% for Gemini 3.8 Flash and 7% for GPT-6 Luna.
- GPT-6 Luna is cheaper at $0.10 / $0.50 per million input/output tokens, against $0.75 / $3.75 for Gemini 3.8 Flash.
- GPT-6 Luna accepts more context: 1.05M tokens versus 1.05M.
Side by side
| Gemini 3.8 Flash | GPT-6 Luna | |
|---|---|---|
| Provider | OpenAI | |
| Noometry Index | 61.8 | 53.3 |
| Released | 2026-09-02 | 2026-09-22 |
| Weights | Proprietary | Proprietary |
| Context window | 1.05M | 1.05M |
| Max output | 66K | 128K |
| Input $ / M tokens | $0.75 | $0.10 |
| Output $ / M tokens | $3.75 | $0.50 |
| Results tracked | 50 | 42 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Gemini 3.8 Flash leads
Gemini 3.8 Flash: 59.2 (#15), GPT-6 Luna: 55.5 (#25)
| Benchmark | Gemini 3.8 Flash | GPT-6 Luna |
|---|---|---|
| DeepSWE | 73.8% | 66.6% |
| FrontierCode | 41.2% | 42.4% |
| LMArena WebDev | 1584 | 1581 |
| SciCode | 56.6% | 54.6% |
| LMArena Coding | 1510 | 1439 |
| ALE-Bench | 1,270 | 1,577 |
| CursorBench | 39.6% | — |
| FrontierSWE | 19.6% | — |
| WeirdML | 84.8% | — |
Agentic & Tool Use Gemini 3.8 Flash leads
Gemini 3.8 Flash: 41.8 (#21), GPT-6 Luna: 33.3 (#54)
| Benchmark | Gemini 3.8 Flash | GPT-6 Luna |
|---|---|---|
| APEX-Agents | 64.3% | 44.3% |
| GDP.pdf | 23.4% | 23% |
| Remote Labor Index | 5.8% | — |
| Vending-Bench 2 | 5,094 | — |
Reasoning Gemini 3.8 Flash leads
Gemini 3.8 Flash: 76.9 (#5), GPT-6 Luna: 48.2 (#41)
| Benchmark | Gemini 3.8 Flash | GPT-6 Luna |
|---|---|---|
| ARC-AGI-2 | 89.2% | 59.3% |
| NYT Connections (extended) | 97.4% | 68.7% |
| ARC-AGI-1 | 98.5% | 86.7% |
| CritPt | 18.3% | 19.4% |
| Chess Puzzles | 61% | 31% |
| LMArena Hard Prompts | 1508 | 1411 |
| Mystery Game Puzzles | 47% | 7% |
| DTBench | 95.7% | 90.1% |
| LMCA | 52.9% | 44.5% |
| Epoch Capabilities Index | 156.71 | 156.28 |
| Surface Evolver Bench | 76.9% | — |
Math GPT-6 Luna leads
Gemini 3.8 Flash: 65.3 (#28), GPT-6 Luna: 76.1 (#15)
| Benchmark | Gemini 3.8 Flash | GPT-6 Luna |
|---|---|---|
| FrontierMath (Tiers 1-3) | 68.4% | 78.9% |
| FrontierMath Tier 4 | 22% | 56.1% |
| OTIS Mock AIME 2024-2025 | 98.9% | 98.9% |
| ProofBench | 48% | 64% |
| LMArena Math | 1528 | 1416 |
Knowledge Gemini 3.8 Flash leads
Gemini 3.8 Flash: 74.8 (#2), GPT-6 Luna: 57.0 (#41)
| Benchmark | Gemini 3.8 Flash | GPT-6 Luna |
|---|---|---|
| GPQA Diamond | 95.4% | 90.5% |
| SimpleQA Verified | 69.7% | 41.4% |
| LMArena Expert | 1524 | 1444 |
| Humanity's Last Exam | 44.5% | — |
Multimodal GPT-6 Luna leads
Gemini 3.8 Flash: 40.7 (#45), GPT-6 Luna: 42.4 (#30)
| Benchmark | Gemini 3.8 Flash | GPT-6 Luna |
|---|---|---|
| LMArena Vision | 1314 | 1217 |
| Blueprint-Bench 2 | 38.6% | 31.2% |
| Furniture Assembly | 31.7% | 44.2% |
Multilingual Gemini 3.8 Flash leads
Gemini 3.8 Flash: 58.0 (#5), GPT-6 Luna: 50.5 (#117)
| Benchmark | Gemini 3.8 Flash | GPT-6 Luna |
|---|---|---|
| LMArena Non-English | 1491 | 1386 |
| LMArena Chinese | 1554 | 1433 |
| LMArena French | 1498 | 1420 |
| LMArena German | 1493 | 1369 |
| LMArena Japanese | 1502 | 1369 |
| LMArena Korean | 1459 | 1360 |
| LMArena Russian | 1515 | 1394 |
| LMArena Spanish | 1485 | 1393 |
Instruction Following Gemini 3.8 Flash leads
Gemini 3.8 Flash: 78.0 (#13), GPT-6 Luna: 74.3 (#99)
| Benchmark | Gemini 3.8 Flash | GPT-6 Luna |
|---|---|---|
| LMArena Instruction Following | 1490 | 1409 |
Long Context Gemini 3.8 Flash leads
Gemini 3.8 Flash: 46.3 (#24), GPT-6 Luna: 43.0 (#111)
| Benchmark | Gemini 3.8 Flash | GPT-6 Luna |
|---|---|---|
| LMArena Longer Query | 1508 | 1409 |
Writing & Preference Gemini 3.8 Flash leads
Gemini 3.8 Flash: 72.2 (#15), GPT-6 Luna: 58.3 (#119)
| Benchmark | Gemini 3.8 Flash | GPT-6 Luna |
|---|---|---|
| LMArena Text | 1499 | 1391 |
| LMArena Creative Writing | 1492 | 1363 |
| LMArena Multi-Turn | 1501 | 1396 |
| EQ-Bench Creative Writing | 1748 | — |
Frequently asked questions
Is Gemini 3.8 Flash better than GPT-6 Luna?
Gemini 3.8 Flash is the stronger model overall, scoring 61.8 to 53.3 on the Noometry Index. GPT-6 Luna costs 7.5× less per token, which makes it the better buy when Gemini 3.8 Flash's lead doesn't matter for your workload.
Which is cheaper, Gemini 3.8 Flash or GPT-6 Luna?
GPT-6 Luna is cheaper. It lists at $0.10 per million input tokens and $0.50 per million output tokens; Gemini 3.8 Flash lists at $0.75 and $3.75.
Is Gemini 3.8 Flash or GPT-6 Luna better for coding?
Gemini 3.8 Flash scores higher on coding benchmarks: 59.2 versus 55.5 in the Noometry coding category.
Which has the bigger context window?
GPT-6 Luna does, with 1.05M tokens against 1.05M.
How many benchmarks do Gemini 3.8 Flash and GPT-6 Luna share?
42 benchmarks have published results for both models. Gemini 3.8 Flash has 50 scored results on Noometry and GPT-6 Luna has 42.