Model comparison
DeepSeek-R1-Distill-Qwen-32B vs GPT-5.6 Luna
GPT-5.6 Luna is the stronger model overall, scoring 54.6 to 35.5 on the Noometry Index.
Last verified . 5 shared benchmarks.
DeepSeekSummary
- They share 5 benchmarks with published results for both. DeepSeek-R1-Distill-Qwen-32B scores higher in 0 categories and GPT-5.6 Luna in 7 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.6 Luna leads 77.7 to 34.5.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 55.6% for DeepSeek-R1-Distill-Qwen-32B and 98.3% for GPT-5.6 Luna.
- DeepSeek-R1-Distill-Qwen-32B has downloadable open weights; the other is API-only.
Side by side
| DeepSeek-R1-Distill-Qwen-32B | GPT-5.6 Luna | |
|---|---|---|
| Provider | DeepSeek | OpenAI |
| Noometry Index | 35.5 | 54.6 |
| Released | 2025-01-20 | 2026-07-09 |
| Weights | Open | Proprietary |
| Context window | — | 1.05M |
| Max output | — | 128K |
| Input $ / M tokens | — | $0.20 |
| Output $ / M tokens | — | $1.20 |
| Results tracked | 14 | 52 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5.6 Luna leads
DeepSeek-R1-Distill-Qwen-32B: 36.1 (#212), GPT-5.6 Luna: 54.5 (#28)
| Benchmark | DeepSeek-R1-Distill-Qwen-32B | GPT-5.6 Luna |
|---|---|---|
| DeepSWE | — | 67.2% |
| FrontierCode | — | 39.8% |
| CursorBench | — | 35.9% |
| LMArena WebDev | — | 1519 |
| SciCode | — | 53.6% |
| WeirdML | — | 60.9% |
| BigCodeBench Instruct | 43.9% | — |
| LiveBench Coding | 33.7% | — |
| LMArena Coding | — | 1466 |
| BigCodeBench Complete | 54.9% | — |
| ALE-Bench | — | 1,667 |
Agentic & Tool Use GPT-5.6 Luna leads
DeepSeek-R1-Distill-Qwen-32B: 28.1 (#94), GPT-5.6 Luna: 34.4 (#45)
| Benchmark | DeepSeek-R1-Distill-Qwen-32B | GPT-5.6 Luna |
|---|---|---|
| BALROG | 19.5% | 45.6% |
| APEX-Agents | — | 43% |
| GDP.pdf | — | 22.7% |
| Vending-Bench 2 | — | 4,095 |
Reasoning GPT-5.6 Luna leads
DeepSeek-R1-Distill-Qwen-32B: 18.2 (#284), GPT-5.6 Luna: 47.6 (#43)
| Benchmark | DeepSeek-R1-Distill-Qwen-32B | GPT-5.6 Luna |
|---|---|---|
| Chess Puzzles | 1% | 40% |
| Epoch Capabilities Index | 137.44 | 156.39 |
| 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% |
| LiveBench Reasoning | 52.3% | — |
| LMArena Hard Prompts | — | 1451 |
| Mystery Game Puzzles | — | 21% |
| DTBench | — | 89.1% |
| LiveBench Data Analysis | 45.4% | — |
| LMCA | — | 48.5% |
| Surface Evolver Bench | — | 61.9% |
| LiveBench | 45.5% | — |
Math GPT-5.6 Luna leads
DeepSeek-R1-Distill-Qwen-32B: 34.5 (#194), GPT-5.6 Luna: 77.7 (#14)
| Benchmark | DeepSeek-R1-Distill-Qwen-32B | GPT-5.6 Luna |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 55.6% | 98.3% |
| FrontierMath (Tiers 1-3) | — | 82.1% |
| FrontierMath Tier 4 | — | 61% |
| ProofBench | — | 60% |
| LiveBench Math | 59.4% | — |
| LMArena Math | — | 1458 |
Knowledge GPT-5.6 Luna leads
DeepSeek-R1-Distill-Qwen-32B: 35.7 (#182), GPT-5.6 Luna: 58.5 (#34)
| Benchmark | DeepSeek-R1-Distill-Qwen-32B | GPT-5.6 Luna |
|---|---|---|
| GPQA Diamond | 64.1% | 91.6% |
| SimpleQA Verified | — | 41% |
| LMArena Expert | — | 1478 |
Multimodal Not comparable
DeepSeek-R1-Distill-Qwen-32B: —, GPT-5.6 Luna: 42.7 (#28)
| Benchmark | DeepSeek-R1-Distill-Qwen-32B | GPT-5.6 Luna |
|---|---|---|
| LMArena Vision | — | 1258 |
| Blueprint-Bench 2 | — | 22.6% |
| Furniture Assembly | — | 42.5% |
| LMArena Document | — | 1457 |
Multilingual Not comparable
DeepSeek-R1-Distill-Qwen-32B: —, GPT-5.6 Luna: 52.8 (#78)
| Benchmark | DeepSeek-R1-Distill-Qwen-32B | GPT-5.6 Luna |
|---|---|---|
| LMArena Non-English | — | 1417 |
| LMArena Chinese | — | 1470 |
| LMArena French | — | 1456 |
| LMArena German | — | 1454 |
| LMArena Japanese | — | 1411 |
| LMArena Korean | — | 1415 |
| LMArena Russian | — | 1428 |
| LMArena Spanish | — | 1448 |
Instruction Following GPT-5.6 Luna leads
DeepSeek-R1-Distill-Qwen-32B: 61.6 (#243), GPT-5.6 Luna: 75.6 (#57)
| Benchmark | DeepSeek-R1-Distill-Qwen-32B | GPT-5.6 Luna |
|---|---|---|
| LiveBench Instruction Following | 55.7% | — |
| LMArena Instruction Following | — | 1437 |
Long Context Not comparable
DeepSeek-R1-Distill-Qwen-32B: —, GPT-5.6 Luna: 43.9 (#82)
| Benchmark | DeepSeek-R1-Distill-Qwen-32B | GPT-5.6 Luna |
|---|---|---|
| LMArena Longer Query | — | 1436 |
Writing & Preference GPT-5.6 Luna leads
DeepSeek-R1-Distill-Qwen-32B: 49.6 (#188), GPT-5.6 Luna: 68.0 (#29)
| Benchmark | DeepSeek-R1-Distill-Qwen-32B | GPT-5.6 Luna |
|---|---|---|
| LMArena Text | — | 1431 |
| LMArena Creative Writing | — | 1396 |
| EQ-Bench Creative Writing | — | 1829 |
| EQ-Bench 4 | — | 1156 |
| LMArena Multi-Turn | — | 1434 |
| LiveBench Language | 26.8% | — |
Frequently asked questions
Is DeepSeek-R1-Distill-Qwen-32B better than GPT-5.6 Luna?
GPT-5.6 Luna is the stronger model overall, scoring 54.6 to 35.5 on the Noometry Index.
Is DeepSeek-R1-Distill-Qwen-32B or GPT-5.6 Luna better for coding?
GPT-5.6 Luna scores higher on coding benchmarks: 54.5 versus 36.1 in the Noometry coding category.
How many benchmarks do DeepSeek-R1-Distill-Qwen-32B and GPT-5.6 Luna share?
5 benchmarks have published results for both models. DeepSeek-R1-Distill-Qwen-32B has 14 scored results on Noometry and GPT-5.6 Luna has 52.