Model comparison
DeepSeek V4 Pro vs GPT-5.6 Luna
DeepSeek V4 Pro and GPT-5.6 Luna score almost the same on the Noometry Index (54.3 vs 54.6), so choose on price, context window or the category you care about most.
Last verified . 43 shared benchmarks.
DeepSeekSummary
- They share 43 benchmarks with published results for both. DeepSeek V4 Pro scores higher in 5 categories and GPT-5.6 Luna in 4 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.6 Luna leads 77.7 to 64.8.
- The biggest single-benchmark swing is FrontierMath Tier 4: 26.8% for DeepSeek V4 Pro and 61% for GPT-5.6 Luna.
- GPT-5.6 Luna is cheaper at $0.20 / $1.20 per million input/output tokens, against $0.66 / $1.98 for DeepSeek V4 Pro.
- GPT-5.6 Luna accepts more context: 1.05M tokens versus 1M.
- DeepSeek V4 Pro has downloadable open weights; the other is API-only.
Side by side
| DeepSeek V4 Pro | GPT-5.6 Luna | |
|---|---|---|
| Provider | DeepSeek | OpenAI |
| Noometry Index | 54.3 | 54.6 |
| Released | 2026-04-24 | 2026-07-09 |
| Weights | Open | Proprietary |
| Context window | 1M | 1.05M |
| Max output | 393K | 128K |
| Input $ / M tokens | $0.66 | $0.20 |
| Output $ / M tokens | $1.98 | $1.20 |
| Results tracked | 48 | 52 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5.6 Luna leads
DeepSeek V4 Pro: 52.4 (#34), GPT-5.6 Luna: 54.5 (#28)
| Benchmark | DeepSeek V4 Pro | GPT-5.6 Luna |
|---|---|---|
| FrontierCode | 28.6% | 39.8% |
| LMArena WebDev | 1582 | 1519 |
| SciCode | 51% | 53.6% |
| WeirdML | 66.2% | 60.9% |
| LMArena Coding | 1470 | 1466 |
| ALE-Bench | 1,403 | 1,667 |
| SWE-bench Verified | 77.6% | — |
| DeepSWE | — | 67.2% |
| CursorBench | — | 35.9% |
Agentic & Tool Use GPT-5.6 Luna leads
DeepSeek V4 Pro: 32.8 (#58), GPT-5.6 Luna: 34.4 (#45)
| Benchmark | DeepSeek V4 Pro | GPT-5.6 Luna |
|---|---|---|
| APEX-Agents | 47.3% | 43% |
| Vending-Bench 2 | 3,285 | 4,095 |
| BALROG | — | 45.6% |
| GDP.pdf | — | 22.7% |
Reasoning DeepSeek V4 Pro leads
DeepSeek V4 Pro: 56.5 (#24), GPT-5.6 Luna: 47.6 (#43)
| Benchmark | DeepSeek V4 Pro | GPT-5.6 Luna |
|---|---|---|
| ARC-AGI-2 | 61.3% | 59.5% |
| Kagi LLM Benchmark | 53.5% | 49.1% |
| NYT Connections (extended) | 91.3% | 69.4% |
| ARC-AGI-1 | 90.5% | 88% |
| CritPt | 18% | 20.6% |
| Chess Puzzles | 47% | 40% |
| LMArena Hard Prompts | 1461 | 1451 |
| Mystery Game Puzzles | 43% | 21% |
| DTBench | 93.9% | 89.1% |
| LMCA | 45.5% | 48.5% |
| Surface Evolver Bench | 40% | 61.9% |
| Epoch Capabilities Index | 155.31 | 156.39 |
| SimpleBench | — | 46.8% |
| ForecastBench | 56.1 | — |
Math GPT-5.6 Luna leads
DeepSeek V4 Pro: 64.8 (#30), GPT-5.6 Luna: 77.7 (#14)
| Benchmark | DeepSeek V4 Pro | GPT-5.6 Luna |
|---|---|---|
| FrontierMath (Tiers 1-3) | 64.6% | 82.1% |
| FrontierMath Tier 4 | 26.8% | 61% |
| OTIS Mock AIME 2024-2025 | 98.6% | 98.3% |
| ProofBench | 50% | 60% |
| LMArena Math | 1455 | 1458 |
| MathArena Final-Answer Competitions | 76.6% | — |
Knowledge Too close to call
DeepSeek V4 Pro: 59.5 (#31), GPT-5.6 Luna: 58.5 (#34)
| Benchmark | DeepSeek V4 Pro | GPT-5.6 Luna |
|---|---|---|
| GPQA Diamond | 91.7% | 91.6% |
| SimpleQA Verified | 52.9% | 41% |
| LMArena Expert | 1464 | 1478 |
| Vectara Hallucination Rate | 8.6% | — |
Multimodal Not comparable
DeepSeek V4 Pro: —, GPT-5.6 Luna: 42.7 (#28)
| Benchmark | DeepSeek V4 Pro | GPT-5.6 Luna |
|---|---|---|
| LMArena Vision | — | 1258 |
| Blueprint-Bench 2 | — | 22.6% |
| Furniture Assembly | — | 42.5% |
| LMArena Document | — | 1457 |
Multilingual DeepSeek V4 Pro leads
DeepSeek V4 Pro: 54.4 (#45), GPT-5.6 Luna: 52.8 (#78)
| Benchmark | DeepSeek V4 Pro | GPT-5.6 Luna |
|---|---|---|
| LMArena Non-English | 1439 | 1417 |
| LMArena Chinese | 1486 | 1470 |
| LMArena French | 1472 | 1456 |
| LMArena German | 1458 | 1454 |
| LMArena Japanese | 1445 | 1411 |
| LMArena Korean | 1447 | 1415 |
| LMArena Russian | 1453 | 1428 |
| LMArena Spanish | 1458 | 1448 |
Instruction Following Too close to call
DeepSeek V4 Pro: 76.1 (#47), GPT-5.6 Luna: 75.6 (#57)
| Benchmark | DeepSeek V4 Pro | GPT-5.6 Luna |
|---|---|---|
| LMArena Instruction Following | 1448 | 1437 |
Long Context DeepSeek V4 Pro leads
DeepSeek V4 Pro: 45.0 (#51), GPT-5.6 Luna: 43.9 (#82)
| Benchmark | DeepSeek V4 Pro | GPT-5.6 Luna |
|---|---|---|
| LMArena Longer Query | 1458 | 1436 |
| CL-bench Life | 13.5% | — |
Writing & Preference GPT-5.6 Luna leads
DeepSeek V4 Pro: 65.5 (#46), GPT-5.6 Luna: 68.0 (#29)
| Benchmark | DeepSeek V4 Pro | GPT-5.6 Luna |
|---|---|---|
| LMArena Text | 1451 | 1431 |
| LMArena Creative Writing | 1446 | 1396 |
| EQ-Bench Creative Writing | 1553 | 1829 |
| EQ-Bench 4 | 1166 | 1156 |
| LMArena Multi-Turn | 1467 | 1434 |
Frequently asked questions
Is DeepSeek V4 Pro better than GPT-5.6 Luna?
DeepSeek V4 Pro and GPT-5.6 Luna score almost the same on the Noometry Index (54.3 vs 54.6), so choose on price, context window or the category you care about most.
Which is cheaper, DeepSeek V4 Pro 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; DeepSeek V4 Pro lists at $0.66 and $1.98.
Is DeepSeek V4 Pro or GPT-5.6 Luna better for coding?
GPT-5.6 Luna scores higher on coding benchmarks: 54.5 versus 52.4 in the Noometry coding category.
Which has the bigger context window?
GPT-5.6 Luna does, with 1.05M tokens against 1M.
How many benchmarks do DeepSeek V4 Pro and GPT-5.6 Luna share?
43 benchmarks have published results for both models. DeepSeek V4 Pro has 48 scored results on Noometry and GPT-5.6 Luna has 52.