Model comparison
DeepSeek-R1 vs Gemini 2.0 Flash (Feb 2025)
DeepSeek-R1 is the stronger model overall, scoring 42.3 to 35.1 on the Noometry Index.
Last verified . 42 shared benchmarks.
DeepSeek
GoogleSummary
- They share 42 benchmarks with published results for both. DeepSeek-R1 scores higher in 8 categories and Gemini 2.0 Flash (Feb 2025) in 1 category; 9 gaps are clear of the uncertainty.
- The widest gap is in coding, where DeepSeek-R1 leads 46.3 to 28.4.
- The biggest single-benchmark swing is Aider Polyglot: 71.4% for DeepSeek-R1 and 38.2% for Gemini 2.0 Flash (Feb 2025).
Side by side
| DeepSeek-R1 | Gemini 2.0 Flash (Feb 2025) | |
|---|---|---|
| Provider | DeepSeek | |
| Noometry Index | 42.3 | 35.1 |
| Released | 2025-01-20 | 2024-12-06 |
| Weights | Proprietary | Proprietary |
| Context window | 164K | — |
| Max output | 64K | — |
| Input $ / M tokens | $0.50 | — |
| Output $ / M tokens | $2.15 | — |
| Results tracked | 52 | 54 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek-R1 leads
DeepSeek-R1: 46.3 (#68), Gemini 2.0 Flash (Feb 2025): 28.4 (#315)
| Benchmark | DeepSeek-R1 | Gemini 2.0 Flash (Feb 2025) |
|---|---|---|
| Aider Polyglot | 71.4% | 38.2% |
| WeirdML | 41.6% | 25.8% |
| LiveBench Coding | 66.7% | 63.4% |
| LMArena Coding | 1427 | 1350 |
| SWE-bench Verified (bash only) | — | 13.5% |
| SciCode | 35.7% | — |
| BigCodeBench Instruct | — | 45.9% |
| BigCodeBench Complete | — | 59.9% |
| CadEval | — | 30% |
| ALE-Bench | 804.12 | — |
| AlgoTune | 1.7 | — |
Agentic & Tool Use DeepSeek-R1 leads
DeepSeek-R1: 30.7 (#75), Gemini 2.0 Flash (Feb 2025): 28.1 (#92)
| Benchmark | DeepSeek-R1 | Gemini 2.0 Flash (Feb 2025) |
|---|---|---|
| TheAgentCompany | — | 11.4% |
| DeepResearch Bench | 35.1% | — |
| BALROG | 34.9% | — |
| METR Time Horizons | 53.8% | — |
Reasoning DeepSeek-R1 leads
DeepSeek-R1: 18.6 (#278), Gemini 2.0 Flash (Feb 2025): 15.2 (#318)
| Benchmark | DeepSeek-R1 | Gemini 2.0 Flash (Feb 2025) |
|---|---|---|
| ARC-AGI-2 | 1.3% | 1.3% |
| SimpleBench | 40.8% | 31.1% |
| Kagi LLM Benchmark | 69.4% | 37.8% |
| LiveBench Reasoning | 83.2% | 78.2% |
| LMArena Hard Prompts | 1416 | 1346 |
| LiveBench Data Analysis | 69.8% | 69.4% |
| Epoch Capabilities Index | 141.29 | 135.36 |
| LiveBench | 71.6% | 66.9% |
| ARC-AGI-1 | 21.2% | — |
| CritPt | 1.1% | — |
| EnigmaEval | — | 1.1% |
| DTBench | — | 63.2% |
| ForecastBench | 60 | — |
Math DeepSeek-R1 leads
DeepSeek-R1: 43.8 (#79), Gemini 2.0 Flash (Feb 2025): 37.9 (#146)
| Benchmark | DeepSeek-R1 | Gemini 2.0 Flash (Feb 2025) |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 66.4% | 57.8% |
| Omni-MATH | 42.4% | 45.9% |
| LiveBench Math | 80.7% | 75.8% |
| LMArena Math | 1400 | 1352 |
| MATH Level 5 | 96.6% | 82.2% |
| FrontierMath (Feb 2025 set) | — | 1.7% |
Knowledge DeepSeek-R1 leads
DeepSeek-R1: 44.5 (#87), Gemini 2.0 Flash (Feb 2025): 32.0 (#213)
| Benchmark | DeepSeek-R1 | Gemini 2.0 Flash (Feb 2025) |
|---|---|---|
| GPQA Diamond | 76.3% | 64.1% |
| MMLU-Pro | 79.3% | 73.7% |
| Confabulations | 12.7% | 12.4% |
| GPQA (HELM) | 66.6% | 55.6% |
| LMArena Expert | 1394 | 1339 |
| Humanity's Last Exam | — | 6.6% |
| Vectara Hallucination Rate | 11.3% | — |
| MMLU | — | 79.7% |
Multimodal Not comparable
DeepSeek-R1: —, Gemini 2.0 Flash (Feb 2025): 36.5 (#79)
| Benchmark | DeepSeek-R1 | Gemini 2.0 Flash (Feb 2025) |
|---|---|---|
| LMArena Vision | — | 1158 |
| GeoBench | — | 77% |
Multilingual DeepSeek-R1 leads
DeepSeek-R1: 52.4 (#85), Gemini 2.0 Flash (Feb 2025): 47.4 (#149)
| Benchmark | DeepSeek-R1 | Gemini 2.0 Flash (Feb 2025) |
|---|---|---|
| LMArena Non-English | 1412 | 1342 |
| LMArena Chinese | 1442 | 1373 |
| LMArena French | 1417 | 1391 |
| LMArena German | 1404 | 1353 |
| LMArena Japanese | 1391 | 1294 |
| LMArena Korean | 1360 | 1313 |
| LMArena Russian | 1423 | 1351 |
| LMArena Spanish | 1411 | 1363 |
Instruction Following Gemini 2.0 Flash (Feb 2025) leads
DeepSeek-R1: 72.0 (#143), Gemini 2.0 Flash (Feb 2025): 74.4 (#97)
| Benchmark | DeepSeek-R1 | Gemini 2.0 Flash (Feb 2025) |
|---|---|---|
| LiveBench Instruction Following | 80.5% | 85.8% |
| IFEval | 78.4% | 84.1% |
| LMArena Instruction Following | 1382 | 1336 |
Long Context DeepSeek-R1 leads
DeepSeek-R1: 45.4 (#36), Gemini 2.0 Flash (Feb 2025): 38.1 (#203)
| Benchmark | DeepSeek-R1 | Gemini 2.0 Flash (Feb 2025) |
|---|---|---|
| Fiction.LiveBench | 75% | 61.1% |
| LMArena Longer Query | 1391 | 1344 |
Writing & Preference DeepSeek-R1 leads
DeepSeek-R1: 61.4 (#88), Gemini 2.0 Flash (Feb 2025): 49.5 (#190)
| Benchmark | DeepSeek-R1 | Gemini 2.0 Flash (Feb 2025) |
|---|---|---|
| LMArena Text | 1428 | 1354 |
| LMArena Creative Writing | 1405 | 1340 |
| Short-Story Creative Writing | 83% | 73.8% |
| EQ-Bench Creative Writing | 1500 | 1128 |
| WildBench | 82.8% | 80% |
| LMArena Multi-Turn | 1405 | 1350 |
| LiveBench Language | 48.5% | 51.3% |
Frequently asked questions
Is DeepSeek-R1 better than Gemini 2.0 Flash (Feb 2025)?
DeepSeek-R1 is the stronger model overall, scoring 42.3 to 35.1 on the Noometry Index.
Is DeepSeek-R1 or Gemini 2.0 Flash (Feb 2025) better for coding?
DeepSeek-R1 scores higher on coding benchmarks: 46.3 versus 28.4 in the Noometry coding category.
How many benchmarks do DeepSeek-R1 and Gemini 2.0 Flash (Feb 2025) share?
42 benchmarks have published results for both models. DeepSeek-R1 has 52 scored results on Noometry and Gemini 2.0 Flash (Feb 2025) has 54.