Model comparison
DeepSeek-R1 vs GPT-5.6 Terra
GPT-5.6 Terra is the stronger model overall, scoring 59.1 to 42.0 on the Noometry Index. DeepSeek-R1 costs 4.9× less per token, which makes it the better buy when GPT-5.6 Terra's lead doesn't matter for your workload.
Last verified . 30 shared benchmarks.
DeepSeekSummary
- They share 30 benchmarks with published results for both. DeepSeek-R1 scores higher in 1 category and GPT-5.6 Terra in 8 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where GPT-5.6 Terra leads 60.6 to 18.7.
- The biggest single-benchmark swing is ARC-AGI-2: 1.3% for DeepSeek-R1 and 83.9% for GPT-5.6 Terra.
- DeepSeek-R1 is cheaper at $0.50 / $2.15 per million input/output tokens, against $2 / $12 for GPT-5.6 Terra.
- GPT-5.6 Terra accepts more context: 1.05M tokens versus 164K.
Side by side
| DeepSeek-R1 | GPT-5.6 Terra | |
|---|---|---|
| Provider | DeepSeek | OpenAI |
| Noometry Index | 42.0 | 59.1 |
| Released | 2025-01-20 | 2026-07-09 |
| Weights | Proprietary | Proprietary |
| Context window | 164K | 1.05M |
| Max output | 64K | 128K |
| Input $ / M tokens | $0.50 | $2 |
| Output $ / M tokens | $2.15 | $12 |
| Results tracked | 52 | 52 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5.6 Terra leads
DeepSeek-R1: 45.9 (#63), GPT-5.6 Terra: 57.1 (#16)
| Benchmark | DeepSeek-R1 | GPT-5.6 Terra |
|---|---|---|
| SciCode | 35.7% | 55% |
| WeirdML | 41.6% | 78.3% |
| LMArena Coding | 1427 | 1484 |
| ALE-Bench | 804.12 | 1,951 |
| DeepSWE | — | 69.6% |
| FrontierCode | — | 41.3% |
| Aider Polyglot | 71.4% | — |
| CursorBench | — | 41.3% |
| LMArena WebDev | — | 1520 |
| LiveBench Coding | 66.7% | — |
| AlgoTune | 1.7 | — |
Agentic & Tool Use GPT-5.6 Terra leads
DeepSeek-R1: 30.4 (#68), GPT-5.6 Terra: 40.2 (#23)
| Benchmark | DeepSeek-R1 | GPT-5.6 Terra |
|---|---|---|
| BALROG | 34.9% | 53.2% |
| APEX-Agents | — | 58.2% |
| DeepResearch Bench | 35.1% | — |
| GDP.pdf | — | 24.7% |
| METR Time Horizons | 53.8% | — |
| Vending-Bench 2 | — | 7,343 |
Reasoning GPT-5.6 Terra leads
DeepSeek-R1: 18.7 (#253), GPT-5.6 Terra: 60.6 (#19)
| Benchmark | DeepSeek-R1 | GPT-5.6 Terra |
|---|---|---|
| ARC-AGI-2 | 1.3% | 83.9% |
| SimpleBench | 40.8% | 48.9% |
| Kagi LLM Benchmark | 69.4% | 51.3% |
| ARC-AGI-1 | 21.2% | 96.5% |
| CritPt | 1.1% | 30% |
| LMArena Hard Prompts | 1415 | 1467 |
| Epoch Capabilities Index | 141.3 | 159.31 |
| NYT Connections (extended) | — | 78.4% |
| Chess Puzzles | — | 54% |
| LiveBench Reasoning | 83.2% | — |
| Mystery Game Puzzles | — | 35% |
| DTBench | — | 93.3% |
| LiveBench Data Analysis | 69.8% | — |
| LMCA | — | 55% |
| Surface Evolver Bench | — | 83.8% |
| ForecastBench | 60 | — |
| LiveBench | 71.6% | — |
Math GPT-5.6 Terra leads
DeepSeek-R1: 42.0 (#70), GPT-5.6 Terra: 81.5 (#10)
| Benchmark | DeepSeek-R1 | GPT-5.6 Terra |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 66.4% | 99.7% |
| LMArena Math | 1401 | 1463 |
| FrontierMath (Tiers 1-3) | — | 86% |
| FrontierMath Tier 4 | — | 70.7% |
| ProofBench | — | 74% |
| Omni-MATH | 42.4% | — |
| LiveBench Math | 80.7% | — |
| MATH Level 5 | 96.6% | — |
Knowledge GPT-5.6 Terra leads
DeepSeek-R1: 44.3 (#81), GPT-5.6 Terra: 61.3 (#27)
| Benchmark | DeepSeek-R1 | GPT-5.6 Terra |
|---|---|---|
| GPQA Diamond | 76.3% | 93.3% |
| LMArena Expert | 1394 | 1492 |
| SimpleQA Verified | — | 43.2% |
| MMLU-Pro | 79.3% | — |
| Confabulations | 12.7% | — |
| Vectara Hallucination Rate | 11.3% | — |
| GPQA (HELM) | 66.6% | — |
Multimodal Not comparable
DeepSeek-R1: —, GPT-5.6 Terra: 47.5 (#10)
| Benchmark | DeepSeek-R1 | GPT-5.6 Terra |
|---|---|---|
| LMArena Vision | — | 1271 |
| Blueprint-Bench 2 | — | 30.8% |
| Furniture Assembly | — | 54.2% |
| LMArena Document | — | 1472 |
Multilingual GPT-5.6 Terra leads
DeepSeek-R1: 52.6 (#80), GPT-5.6 Terra: 54.5 (#42)
| Benchmark | DeepSeek-R1 | GPT-5.6 Terra |
|---|---|---|
| LMArena Non-English | 1412 | 1439 |
| LMArena Chinese | 1443 | 1514 |
| LMArena French | 1415 | 1476 |
| LMArena German | 1405 | 1462 |
| LMArena Japanese | 1394 | 1446 |
| LMArena Korean | 1359 | 1420 |
| LMArena Russian | 1423 | 1450 |
| LMArena Spanish | 1410 | 1455 |
Instruction Following GPT-5.6 Terra leads
DeepSeek-R1: 72.2 (#138), GPT-5.6 Terra: 76.6 (#38)
| Benchmark | DeepSeek-R1 | GPT-5.6 Terra |
|---|---|---|
| LMArena Instruction Following | 1382 | 1453 |
| LiveBench Instruction Following | 80.5% | — |
| IFEval | 78.4% | — |
Long Context DeepSeek-R1 leads
DeepSeek-R1: 45.4 (#36), GPT-5.6 Terra: 44.4 (#64)
| Benchmark | DeepSeek-R1 | GPT-5.6 Terra |
|---|---|---|
| LMArena Longer Query | 1391 | 1452 |
| Fiction.LiveBench | 75% | — |
Writing & Preference GPT-5.6 Terra leads
DeepSeek-R1: 61.5 (#84), GPT-5.6 Terra: 70.4 (#21)
| Benchmark | DeepSeek-R1 | GPT-5.6 Terra |
|---|---|---|
| LMArena Text | 1428 | 1446 |
| LMArena Creative Writing | 1406 | 1411 |
| EQ-Bench Creative Writing | 1500 | 1855 |
| LMArena Multi-Turn | 1404 | 1451 |
| Short-Story Creative Writing | 83% | — |
| WildBench | 82.8% | — |
| EQ-Bench 4 | — | 1234 |
| LiveBench Language | 48.5% | — |
Frequently asked questions
Is DeepSeek-R1 better than GPT-5.6 Terra?
GPT-5.6 Terra is the stronger model overall, scoring 59.1 to 42.0 on the Noometry Index. DeepSeek-R1 costs 4.9× less per token, which makes it the better buy when GPT-5.6 Terra's lead doesn't matter for your workload.
Which is cheaper, DeepSeek-R1 or GPT-5.6 Terra?
DeepSeek-R1 is cheaper. It lists at $0.50 per million input tokens and $2.15 per million output tokens; GPT-5.6 Terra lists at $2 and $12.
Is DeepSeek-R1 or GPT-5.6 Terra better for coding?
GPT-5.6 Terra scores higher on coding benchmarks: 57.1 versus 45.9 in the Noometry coding category.
Which has the bigger context window?
GPT-5.6 Terra does, with 1.05M tokens against 164K.
How many benchmarks do DeepSeek-R1 and GPT-5.6 Terra share?
30 benchmarks have published results for both models. DeepSeek-R1 has 52 scored results on Noometry and GPT-5.6 Terra has 52.