Gemini 3.1 Pro Preview vs GPT-5.6 Luna
GPT-5.6 Luna leads Gemini 3.1 Pro Preview on coding, 58.0 to 32.3 on the Vibe Coding Index, and GPT-5.6 Luna costs 10.0x less per output token ($1.20 vs $12.00 per million). Updated October 22, 2026; 106 benchmarks have a result for both.
Gemini 3.1 Pro Preview or GPT-5.6 Luna: which to pick
Pick GPT-5.6 Luna for coding: it scores 58.0 on the Vibe Coding Index against 32.3 for Gemini 3.1 Pro Preview.
Pick Gemini 3.1 Pro Preview for work measured by SciCode or WeirdML, where it leads.
- Shared results
- 106
- Gemini 3.1 Pro Preview only
- 27
- GPT-5.6 Luna only
- 13
- Like-for-like categories
- 8 of 9
Only benchmarks that tested both models count toward a pick; a result only one model has is shown but not compared.
Which one for your work
Each card is decided by one number both models carry. Where only one has it, the card says so and makes no call.
Coding
Vibe Coding Index, six public coding benchmarks
GPT-5.6 Luna
GPT-5.6 Luna scores 58.0 against 32.3.
4 and 6 benchmarks behind the scores
Intelligence
AA Intelligence Index
GPT-5.6 Luna
GPT-5.6 Luna leads, 37.3 against 29.7.
Same board, both measured
Agentic work
AA Agentic Index
GPT-5.6 Luna
GPT-5.6 Luna leads, 42.1 against 8.2.
Same board, both measured
Reasoning
GPQA Diamond
Gemini 3.1 Pro Preview
Gemini 3.1 Pro Preview leads, 95.5% against 91.7%.
Same board, both measured
Math
AIME
No call
Only Gemini 3.1 Pro Preview has a AIME result, 98.1%.
Directional only
Knowledge
MMLU Pro
Gemini 3.1 Pro Preview
Gemini 3.1 Pro Preview leads, 91.0% against 86.0%.
Same board, both measured
Web design
Design Arena: websites
GPT-5.6 Luna
GPT-5.6 Luna leads, 1255 against 1233.
Same board, both measured
Long documents
Listed context window
GPT-5.6 Luna
GPT-5.6 Luna takes 1.05M tokens against 1.05M.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
GPT-5.6 Luna
Gemini 3.1 Pro Preview $0.0080, GPT-5.6 Luna $0.0008 per turn.
List prices
Cost per solved task
One SWE-bench Verified task, solved
GPT-5.6 Luna
Gemini 3.1 Pro Preview $0.99, GPT-5.6 Luna $0.04.
Published attempt cost and pass rate
At a glance
- Coding index
- Gemini 3.1 Pro Preview 32.3 from 4 benchmarks, GPT-5.6 Luna 58.0 from 6 benchmarks.
- Widest category
- Math, 16.5 points apart on average over 6 shared benchmarks.
- Widest single gap
- Vibe Code Bench (Vals AI): Gemini 3.1 Pro Preview 32.0% against GPT-5.6 Luna 77.1%, 45.0 points apart on a 0 to 100 scale.
- Output price
- Gemini 3.1 Pro Preview $12.00 against GPT-5.6 Luna $1.20 per million output tokens, 10.0x apart.
- Context window
- Gemini 3.1 Pro Preview takes 1.05M tokens, GPT-5.6 Luna 1.05M.
Benchmark by benchmark
First the score: boards each model leads, per category, where both have a result. Then every board, a category to a fold; bars are each result on a 0 to 100 scale, so compare the two in a row, not one row with another.
| Category | Gemini 3.1 Pro Preview | Bars | GPT-5.6 Luna | Ties |
|---|---|---|---|---|
| Coding | 5 | 13 | 1 tie | |
| Intelligence | 0 | 1 | ||
| Agentic work | 3 | 9 | ||
| Reasoning | 16 | 0 | ||
| Math | 2 | 4 | ||
| Knowledge | 11 | 1 | ||
| Web design | 4 | 3 | 1 tie | |
| General | 26 | 6 | ||
| All shared boards | 67 | 37 | 2 ties |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index513
| Benchmark | Gemini 3.1 Pro Preview | GPT-5.6 Luna | Leader |
|---|---|---|---|
| LiveBench coding | Not measured | 73.6% | Only GPT-5.6 Luna measured |
| LMArena WebDev | 1447 | 1518 | GPT-5.6 Luna |
| AA Coding Index | 68.8 | 71.4 | GPT-5.6 Luna |
| LiveCodeBench (Vals AI) | 88.5% | Not measured | Only Gemini 3.1 Pro Preview measured |
| SWE-bench Verified (Vals AI) | 78.8% | 93.0% | GPT-5.6 Luna |
| Terminal-Bench 4 (Vals AI) | 2.5% | 11.6% | GPT-5.6 Luna |
| Vibe Code Bench (Vals AI) | 32.0% | 77.1% | GPT-5.6 Luna |
| Code migration (Vals AI) | 17.3% | 44.5% | GPT-5.6 Luna |
| ProgramBench (Vals AI) | 0.0% | 0.0% | Tie |
| IOI (Vals AI) | 51.8% | 61.8% | GPT-5.6 Luna |
| Terminal-Bench 2.1 (Vals AI) | 70.8% | 79.0% | GPT-5.6 Luna |
| Vibe Code Bench 1-100 (Vals AI) | 6.7% | 22.6% | GPT-5.6 Luna |
| SkillsBench (Vals AI) | Not measured | 60.5% | Only GPT-5.6 Luna measured |
| CyberBench (Vals AI) | Not measured | 73.6% | Only GPT-5.6 Luna measured |
| SciCode | 58.7% | 53.6% | Gemini 3.1 Pro Preview |
| Terminal-Bench Hard (Artificial Analysis run) | 53.8% | Not measured | Only Gemini 3.1 Pro Preview measured |
| Terminal-Bench | 80.2% | Not measured | Only Gemini 3.1 Pro Preview measured |
| FrontierCode | Not measured | 39.8% | Only GPT-5.6 Luna measured |
| GSO-Bench | 22.6% | Not measured | Only Gemini 3.1 Pro Preview measured |
| WeirdML | 72.1% | 60.9% | Gemini 3.1 Pro Preview |
| MirrorCode | 8.9% | Not measured | Only Gemini 3.1 Pro Preview measured |
| ALE-Bench | 1160.6 | 1667.4 | GPT-5.6 Luna |
| LMArena Text: coding | 1484 | 1466 | Gemini 3.1 Pro Preview |
| LMArena Text: software and IT | 1487 | 1463 | Gemini 3.1 Pro Preview |
| LMArena WebDev: React | 1438 | 1508 | GPT-5.6 Luna |
| LMArena WebDev: HTML | 1490 | 1520 | GPT-5.6 Luna |
| LMArena WebDev: from an image | 1481 | 1517 | GPT-5.6 Luna |
| SciCode | 14.3% | 6.2% | Gemini 3.1 Pro Preview |
| Neither model has a result on the other 11 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index01
| Benchmark | Gemini 3.1 Pro Preview | GPT-5.6 Luna | Leader |
|---|---|---|---|
| AA Intelligence Index | 29.7 | 37.3 | GPT-5.6 Luna |
| Epoch Capabilities Index | Not measured | 156.5 | Only GPT-5.6 Luna measured |
Agentic workLike-for-likePublic boards, shown as published, kept out of the index39
| Benchmark | Gemini 3.1 Pro Preview | GPT-5.6 Luna | Leader |
|---|---|---|---|
| AA Agentic Index | 8.2 | 42.1 | GPT-5.6 Luna |
| Design Arena: agents building web apps | 1125 | Not measured | Only Gemini 3.1 Pro Preview measured |
| Design Arena: agents building full-stack apps | 1068 | Not measured | Only Gemini 3.1 Pro Preview measured |
| Design Arena: agents building mobile apps | 1094 | Not measured | Only Gemini 3.1 Pro Preview measured |
| Design Arena: agents building Android apps | 1049 | Not measured | Only Gemini 3.1 Pro Preview measured |
| Design Arena: agents building Godot games | 1236 | Not measured | Only Gemini 3.1 Pro Preview measured |
| Design Arena: agents building games | 1113 | Not measured | Only Gemini 3.1 Pro Preview measured |
| Design Arena: HTML slides | 1103 | Not measured | Only Gemini 3.1 Pro Preview measured |
| Design Arena: PowerPoint slides | 1041 | Not measured | Only Gemini 3.1 Pro Preview measured |
| LiveBench research | Not measured | 51.0% | Only GPT-5.6 Luna measured |
| LiveBench browser | Not measured | 68.0% | Only GPT-5.6 Luna measured |
| LiveBench long agentic | Not measured | 30.4% | Only GPT-5.6 Luna measured |
| SREBench (Vals AI) | 0.0% | 1.5% | GPT-5.6 Luna |
| Finance Agent (Vals AI) | 43.0% | 55.0% | GPT-5.6 Luna |
| Tax Agent Bench (Vals AI) | 9.3% | 24.8% | GPT-5.6 Luna |
| Legal Research Bench (Vals AI) | 20.7% | 36.5% | GPT-5.6 Luna |
| Excel Modeling Benchmark (Vals AI) | 52.6% | 67.1% | GPT-5.6 Luna |
| BioMysteryBench (Vals AI) | 60.0% | 61.5% | GPT-5.6 Luna |
| Terminal-Bench-Science (Vals AI) | 1.4% | 0.0% | Gemini 3.1 Pro Preview |
| Vals Science Index | 27.4% | 25.3% | Gemini 3.1 Pro Preview |
| τ²-Bench Telecom | 95.6% | Not measured | Only Gemini 3.1 Pro Preview measured |
| APEX-Agents | 35.1% | 43.0% | GPT-5.6 Luna |
| METR time horizons | 77.0% | Not measured | Only Gemini 3.1 Pro Preview measured |
| BALROG | 57.0% | 45.6% | Gemini 3.1 Pro Preview |
| DeepResearch Bench | 47.8% | Not measured | Only Gemini 3.1 Pro Preview measured |
| PostTrainBench | 22.0% | Not measured | Only Gemini 3.1 Pro Preview measured |
| Vending-Bench 2 | $911 | $4,095 | GPT-5.6 Luna |
| Neither model has a result on the other 14 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index160
| Benchmark | Gemini 3.1 Pro Preview | GPT-5.6 Luna | Leader |
|---|---|---|---|
| GPQA Diamond | 95.5% | 91.7% | Gemini 3.1 Pro Preview |
| LiveBench reasoning | Not measured | 85.6% | Only GPT-5.6 Luna measured |
| MysteryMechanism (Vals AI) | 20.7% | 14.4% | Gemini 3.1 Pro Preview |
| MMMU Pro (Vals AI) | 88.2% | 85.0% | Gemini 3.1 Pro Preview |
| GPQA Diamond (Artificial Analysis run) | 94.1% | 91.1% | Gemini 3.1 Pro Preview |
| Humanity's Last Exam (Artificial Analysis run) | 47.0% | 39.5% | Gemini 3.1 Pro Preview |
| GPQA Diamond (Epoch AI run) | 94.4% | 91.6% | Gemini 3.1 Pro Preview |
| ARC-AGI | 98.0% | 88.0% | Gemini 3.1 Pro Preview |
| ARC-AGI-2 | 77.1% | 59.5% | Gemini 3.1 Pro Preview |
| SimpleBench | 79.6% | 46.8% | Gemini 3.1 Pro Preview |
| Chess puzzles | 51.0% | 43.0% | Gemini 3.1 Pro Preview |
| Mystery game puzzles | 34.0% | 21.0% | Gemini 3.1 Pro Preview |
| EBR-bench | 14.3% | Not measured | Only Gemini 3.1 Pro Preview measured |
| LMArena Text: hard prompts | 1485 | 1451 | Gemini 3.1 Pro Preview |
| LMArena Text: expert | 1483 | 1478 | Gemini 3.1 Pro Preview |
| LMArena Text: hard prompts in English | 1480 | 1460 | Gemini 3.1 Pro Preview |
| LMArena Vision: diagrams | 1311 | 1273 | Gemini 3.1 Pro Preview |
| LMArena Vision: homework | 1316 | 1282 | Gemini 3.1 Pro Preview |
| Kagi LLM Benchmark | Not measured | 49.1% | Only GPT-5.6 Luna measured |
| Neither model has a result on the other 3 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index24
| Benchmark | Gemini 3.1 Pro Preview | GPT-5.6 Luna | Leader |
|---|---|---|---|
| AIME | 98.1% | Not measured | Only Gemini 3.1 Pro Preview measured |
| OTIS Mock AIME | 95.6% | 98.3% | GPT-5.6 Luna |
| FrontierMath, tiers 1 to 3 | 59.6% | 82.1% | GPT-5.6 Luna |
| FrontierMath, tier 4 | 26.8% | 61.0% | GPT-5.6 Luna |
| ProofBench | 26.0% | 60.0% | GPT-5.6 Luna |
| LMArena Text: math | 1486 | 1459 | Gemini 3.1 Pro Preview |
| LMArena Text: mathematical work | 1478 | 1470 | Gemini 3.1 Pro Preview |
| Neither model has a result on the other 6 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index111
| Benchmark | Gemini 3.1 Pro Preview | GPT-5.6 Luna | Leader |
|---|---|---|---|
| MMLU Pro | 91.0% | 86.0% | Gemini 3.1 Pro Preview |
| LegalBench (Vals AI) | 87.4% | 84.0% | Gemini 3.1 Pro Preview |
| CaseLaw (Vals AI) | 64.8% | Not measured | Only Gemini 3.1 Pro Preview measured |
| CorpFin (Vals AI) | 64.5% | 64.2% | Gemini 3.1 Pro Preview |
| TaxEval (Vals AI) | 72.9% | 76.2% | GPT-5.6 Luna |
| MortgageTax (Vals AI) | 69.4% | 67.3% | Gemini 3.1 Pro Preview |
| MedQA (Vals AI) | 96.4% | Not measured | Only Gemini 3.1 Pro Preview measured |
| MedCode (Vals AI) | 59.1% | 42.4% | Gemini 3.1 Pro Preview |
| SimpleQA Verified | 73.5% | 41.0% | Gemini 3.1 Pro Preview |
| LMArena Text: science | 1497 | 1433 | Gemini 3.1 Pro Preview |
| LMArena Text: medicine and healthcare | 1486 | 1399 | Gemini 3.1 Pro Preview |
| LMArena Text: legal and government | 1487 | 1446 | Gemini 3.1 Pro Preview |
| LMArena Text: business and finance | 1459 | 1439 | Gemini 3.1 Pro Preview |
| FACTS Benchmark Suite | 67.4% | 35.0% | Gemini 3.1 Pro Preview |
| Neither model has a result on the other 4 boards here | |||
Web designLike-for-likePublic boards, shown as published, kept out of the index43
| Benchmark | Gemini 3.1 Pro Preview | GPT-5.6 Luna | Leader |
|---|---|---|---|
| Design Arena: websites | 1233 | 1255 | GPT-5.6 Luna |
| Design Arena: UI components | 1224 | 1248 | GPT-5.6 Luna |
| Design Arena: game dev | 1202 | 1194 | Gemini 3.1 Pro Preview |
| Design Arena: data viz | 1209 | 1211 | GPT-5.6 Luna |
| Design Arena: 3D | 1234 | 1121 | Gemini 3.1 Pro Preview |
| Design Arena: all code categories | 1231 | 1231 | Tie |
| Design Arena: SVG | 1285 | 1194 | Gemini 3.1 Pro Preview |
| Design Arena: ASCII art | 1270 | 1142 | Gemini 3.1 Pro Preview |
| LiveBench design | Not measured | 76.0% | Only GPT-5.6 Luna measured |
GeneralLike-for-likePublic boards, shown as published, kept out of the index266
| Benchmark | Gemini 3.1 Pro Preview | GPT-5.6 Luna | Leader |
|---|---|---|---|
| LiveBench data analysis | Not measured | 78.0% | Only GPT-5.6 Luna measured |
| Public Benefits Bench (Vals AI) | 53.8% | 61.2% | GPT-5.6 Luna |
| SAGE (Vals AI) | 48.7% | 44.2% | Gemini 3.1 Pro Preview |
| MedScribe (Vals AI) | 76.1% | 84.4% | GPT-5.6 Luna |
| Vals Multimodal Index | 56.1% | 69.1% | GPT-5.6 Luna |
| IFBench | 77.1% | Not measured | Only Gemini 3.1 Pro Preview measured |
| Long context reasoning (AA-LCR) | 82.0% | 83.7% | GPT-5.6 Luna |
| DTBench | 97.1% | 89.1% | Gemini 3.1 Pro Preview |
| LMCA | 53.8% | 48.5% | Gemini 3.1 Pro Preview |
| CL-bench | 20.8% | Not measured | Only Gemini 3.1 Pro Preview measured |
| Surface Evolver Bench | Not measured | 61.9% | Only GPT-5.6 Luna measured |
| Furniture assembly | 26.7% | 42.5% | GPT-5.6 Luna |
| ExploitBench | 26.1% | Not measured | Only Gemini 3.1 Pro Preview measured |
| CL-bench Life | 16.9% | Not measured | Only Gemini 3.1 Pro Preview measured |
| LMArena Text | 1480 | 1431 | Gemini 3.1 Pro Preview |
| LMArena Text: instruction following | 1466 | 1437 | Gemini 3.1 Pro Preview |
| LMArena Text: creative writing | 1482 | 1396 | Gemini 3.1 Pro Preview |
| LMArena Vision | 1296 | 1257 | Gemini 3.1 Pro Preview |
| LMArena Text: English | 1478 | 1441 | Gemini 3.1 Pro Preview |
| LMArena Text: non-English | 1476 | 1417 | Gemini 3.1 Pro Preview |
| LMArena Text: Chinese | 1527 | 1470 | Gemini 3.1 Pro Preview |
| LMArena Text: German | 1490 | 1455 | Gemini 3.1 Pro Preview |
| LMArena Text: Japanese | 1491 | 1415 | Gemini 3.1 Pro Preview |
| LMArena Text: Korean | 1455 | 1415 | Gemini 3.1 Pro Preview |
| LMArena Text: French | 1487 | 1458 | Gemini 3.1 Pro Preview |
| LMArena Text: Spanish | 1481 | 1446 | Gemini 3.1 Pro Preview |
| LMArena Text: Russian | 1497 | 1427 | Gemini 3.1 Pro Preview |
| LMArena Text: Polish | 1498 | 1405 | Gemini 3.1 Pro Preview |
| LMArena Text: multi-turn | 1487 | 1434 | Gemini 3.1 Pro Preview |
| LMArena Text: longer queries | 1482 | 1436 | Gemini 3.1 Pro Preview |
| LMArena Text: writing and language | 1481 | 1419 | Gemini 3.1 Pro Preview |
| LMArena Text: entertainment and media | 1465 | 1402 | Gemini 3.1 Pro Preview |
| LMArena Vision: English | 1287 | 1263 | Gemini 3.1 Pro Preview |
| LMArena Vision: Chinese | 1364 | 1294 | Gemini 3.1 Pro Preview |
| LMArena Vision: reading text | 1304 | 1264 | Gemini 3.1 Pro Preview |
| LMArena Vision: humor | 1322 | 1190 | Gemini 3.1 Pro Preview |
| LMArena Vision: captioning | 1345 | Not measured | Only Gemini 3.1 Pro Preview measured |
| LMArena Vision: recognizing things | 1330 | Not measured | Only Gemini 3.1 Pro Preview measured |
| LMArena Vision: creative writing | 1324 | 1257 | Gemini 3.1 Pro Preview |
| LMArena Document | 1444 | 1457 | GPT-5.6 Luna |
| Neither model has a result on the other 6 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Gemini 3.1 Pro Preview | GPT-5.6 Luna | Leader |
|---|---|---|---|
| Neither model has a result on 5 boards here | |||
What each workload costs
Three fixed token mixes at each model's listed price, so per-token rates become a bill.
Chat turn
1K input + 500 output tokens
GPT-5.6 Luna costs less
Repository review
50K input + 3K output tokens
GPT-5.6 Luna costs less
Agent loop
200K input + 10K output tokens
GPT-5.6 Luna costs less
Listed OpenRouter rates, without cache or batch discounts. A workload that does not fit a model's context window would need splitting.
Specification differences
The leading value in a row is marked; a row with a missing value names no leader. Our two rows are marked in red.
| Specification | Google Gemini 3.1 Pro Preview | OpenAI GPT-5.6 Luna |
|---|---|---|
| Maker | OpenAI | |
| Weights | No public weights listed | No public weights listed |
| Released | February 19, 2026 | July 9, 2026 |
| Context window | 1.05M | 1.05M |
| Input $/Mtok | $2.00 | $0.20 |
| Output $/Mtok | $12.00 | $1.20 |
| Coding index | 32.3 | 58.0 |
| Independent measurements behind it | 4 | 6 |
| Cost per solved task | $0.99 | $0.04 |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is Gemini 3.1 Pro Preview better than GPT-5.6 Luna for coding?
Pick GPT-5.6 Luna for coding: it scores 58.0 on the Vibe Coding Index against 32.3 for Gemini 3.1 Pro Preview. Pick Gemini 3.1 Pro Preview for work measured by SciCode or WeirdML, where it leads.
Which is cheaper, Gemini 3.1 Pro Preview or GPT-5.6 Luna?
Gemini 3.1 Pro Preview costs $2.00 per million input tokens and $12.00 per million output tokens; GPT-5.6 Luna costs $0.20 and $1.20. Per solved SWE-bench Verified task, Gemini 3.1 Pro Preview costs $0.99 and GPT-5.6 Luna $0.04.
Which has the bigger context window, Gemini 3.1 Pro Preview or GPT-5.6 Luna?
GPT-5.6 Luna, with 1.05M tokens against 1.05M.
How many benchmarks is the Gemini 3.1 Pro Preview vs GPT-5.6 Luna comparison based on?
106 benchmarks have a result for both models as of October 22, 2026. Gemini 3.1 Pro Preview's coding index rests on 4 independent benchmarks and GPT-5.6 Luna's on 6. Where only one model has a result, the page says so instead of naming a winner.