# DeepSeek-R1 vs GPT-6 Sol

> GPT-6 Sol is the stronger model overall, scoring 61.8 to 42.3 on the Noometry Index. DeepSeek-R1 costs 4.4× less per token, which makes it the better buy when GPT-6 Sol's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/deepseek-r1-vs-gpt-6-sol
- Last updated: 2026-10-11
- Shared benchmarks: 27

## Summary

- They share 27 benchmarks with published results for both. DeepSeek-R1 scores higher in 2 categories and GPT-6 Sol in 7 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where GPT-6 Sol leads 74.0 to 18.6.
- The biggest single-benchmark swing is ARC-AGI-2: 1.3% for DeepSeek-R1 and 89.6% for GPT-6 Sol.
- DeepSeek-R1 is cheaper at $0.50 / $2.15 per million input/output tokens, against $2 / $10 for GPT-6 Sol.
- GPT-6 Sol accepts more context: 1.05M tokens versus 164K.

## Snapshot

| | DeepSeek-R1 | GPT-6 Sol |
|---|---|---|
| Provider | DeepSeek | OpenAI |
| Noometry Index | 42.3 | 61.8 |
| Rank | 115 | 12 |
| Context | 164K | 1.05M |
| Input $/M | $0.50 | $2 |
| Output $/M | $2.15 | $10 |
| Weights | Proprietary | Proprietary |

## Coding

- DeepSeek-R1: 46.3 (#68)
- GPT-6 Sol: 60.1 (#11)

| Benchmark | DeepSeek-R1 | GPT-6 Sol |
|---|---|---|
| SciCode | 35.7% | 57.6% |
| LMArena Coding | 1427 | 1447 |
| ALE-Bench | 804.12 | 2,462 |
| DeepSWE | — | 68.8% |
| FrontierCode | — | 49.3% |
| Aider Polyglot | 71.4% | — |
| LMArena WebDev | — | 1688 |
| WeirdML | 41.6% | — |
| LiveBench Coding | 66.7% | — |
| AlgoTune | 1.7 | — |

## Agentic & Tool Use

- DeepSeek-R1: 30.7 (#75)
- GPT-6 Sol: 37.2 (#36)

| Benchmark | DeepSeek-R1 | GPT-6 Sol |
|---|---|---|
| APEX-Agents | — | 54.3% |
| DeepResearch Bench | 35.1% | — |
| BALROG | 34.9% | — |
| GDP.pdf | — | 26.4% |
| METR Time Horizons | 53.8% | — |
| Vending-Bench 2 | — | 14,428 |

## Reasoning

- DeepSeek-R1: 18.6 (#278)
- GPT-6 Sol: 74.0 (#9)

| Benchmark | DeepSeek-R1 | GPT-6 Sol |
|---|---|---|
| ARC-AGI-2 | 1.3% | 89.6% |
| ARC-AGI-1 | 21.2% | 95.5% |
| CritPt | 1.1% | 30.9% |
| LMArena Hard Prompts | 1416 | 1418 |
| Epoch Capabilities Index | 141.29 | 162.72 |
| SimpleBench | 40.8% | — |
| Kagi LLM Benchmark | 69.4% | — |
| NYT Connections (extended) | — | 90.1% |
| EBR-Bench | — | 53.3% |
| LiveBench Reasoning | 83.2% | — |
| Mystery Game Puzzles | — | 56% |
| DTBench | — | 97.3% |
| LiveBench Data Analysis | 69.8% | — |
| LMCA | — | 59.1% |
| ForecastBench | 60 | — |
| LiveBench | 71.6% | — |

## Math

- DeepSeek-R1: 43.8 (#79)
- GPT-6 Sol: 87.2 (#7)

| Benchmark | DeepSeek-R1 | GPT-6 Sol |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 66.4% | 100% |
| LMArena Math | 1400 | 1402 |
| FrontierMath (Tiers 1-3) | — | 89.8% |
| FrontierMath Tier 4 | — | 90% |
| ProofBench | — | 83% |
| Omni-MATH | 42.4% | — |
| LiveBench Math | 80.7% | — |
| MATH Level 5 | 96.6% | — |

## Knowledge

- DeepSeek-R1: 44.5 (#87)
- GPT-6 Sol: 64.8 (#15)

| Benchmark | DeepSeek-R1 | GPT-6 Sol |
|---|---|---|
| GPQA Diamond | 76.3% | 94.3% |
| Vectara Hallucination Rate | 11.3% | 6.5% |
| LMArena Expert | 1394 | 1439 |
| SimpleQA Verified | — | 60.7% |
| MMLU-Pro | 79.3% | — |
| Confabulations | 12.7% | — |
| GPQA (HELM) | 66.6% | — |

## Multimodal

- DeepSeek-R1: —
- GPT-6 Sol: 47.6 (#10)

| Benchmark | DeepSeek-R1 | GPT-6 Sol |
|---|---|---|
| LMArena Vision | — | 1245 |
| Blueprint-Bench 2 | — | 36.9% |
| Furniture Assembly | — | 58.3% |

## Multilingual

- DeepSeek-R1: 52.4 (#85)
- GPT-6 Sol: 50.5 (#118)

| Benchmark | DeepSeek-R1 | GPT-6 Sol |
|---|---|---|
| LMArena Non-English | 1412 | 1385 |
| LMArena Chinese | 1442 | 1405 |
| LMArena French | 1417 | 1410 |
| LMArena German | 1404 | 1390 |
| LMArena Japanese | 1391 | 1385 |
| LMArena Korean | 1360 | 1341 |
| LMArena Russian | 1423 | 1401 |
| LMArena Spanish | 1411 | 1384 |

## Instruction Following

- DeepSeek-R1: 72.0 (#143)
- GPT-6 Sol: 74.5 (#94)

| Benchmark | DeepSeek-R1 | GPT-6 Sol |
|---|---|---|
| LMArena Instruction Following | 1382 | 1412 |
| LiveBench Instruction Following | 80.5% | — |
| IFEval | 78.4% | — |

## Long Context

- DeepSeek-R1: 45.4 (#36)
- GPT-6 Sol: 43.1 (#108)

| Benchmark | DeepSeek-R1 | GPT-6 Sol |
|---|---|---|
| LMArena Longer Query | 1391 | 1411 |
| Fiction.LiveBench | 75% | — |

## Writing & Preference

- DeepSeek-R1: 61.4 (#88)
- GPT-6 Sol: 71.9 (#18)

| Benchmark | DeepSeek-R1 | GPT-6 Sol |
|---|---|---|
| LMArena Text | 1428 | 1395 |
| LMArena Creative Writing | 1405 | 1378 |
| EQ-Bench Creative Writing | 1500 | 2125 |
| LMArena Multi-Turn | 1405 | 1412 |
| Short-Story Creative Writing | 83% | — |
| WildBench | 82.8% | — |
| LiveBench Language | 48.5% | — |

## FAQ

### Is DeepSeek-R1 better than GPT-6 Sol?

GPT-6 Sol is the stronger model overall, scoring 61.8 to 42.3 on the Noometry Index. DeepSeek-R1 costs 4.4× less per token, which makes it the better buy when GPT-6 Sol's lead doesn't matter for your workload.

### Which is cheaper, DeepSeek-R1 or GPT-6 Sol?

DeepSeek-R1 is cheaper. It lists at $0.50 per million input tokens and $2.15 per million output tokens; GPT-6 Sol lists at $2 and $10.

### Is DeepSeek-R1 or GPT-6 Sol better for coding?

GPT-6 Sol scores higher on coding benchmarks: 60.1 versus 46.3 in the Noometry coding category.

### Which has the bigger context window?

GPT-6 Sol does, with 1.05M tokens against 164K.

### How many benchmarks do DeepSeek-R1 and GPT-6 Sol share?

27 benchmarks have published results for both models. DeepSeek-R1 has 52 scored results on Noometry and GPT-6 Sol has 45.