Model comparison
DeepSeek-R1 vs Grok-3 mini
DeepSeek-R1 is the stronger model overall, scoring 42.0 to 41.0 on the Noometry Index.
Last verified . 34 shared benchmarks.
DeepSeekSummary
- They share 34 benchmarks with published results for both. DeepSeek-R1 scores higher in 6 categories and Grok-3 mini in 2 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where DeepSeek-R1 leads 61.5 to 52.7.
- The biggest single-benchmark swing is Aider Polyglot: 71.4% for DeepSeek-R1 and 49.3% for Grok-3 mini.
Side by side
| DeepSeek-R1 | Grok-3 mini | |
|---|---|---|
| Provider | DeepSeek | xAI |
| Noometry Index | 42.0 | 41.0 |
| Released | 2025-01-20 | 2025-04-09 |
| Weights | Proprietary | Proprietary |
| Context window | 164K | — |
| Max output | 64K | — |
| Input $ / M tokens | $0.50 | — |
| Output $ / M tokens | $2.15 | — |
| Results tracked | 52 | 35 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek-R1 leads
DeepSeek-R1: 45.9 (#63), Grok-3 mini: 40.4 (#126)
| Benchmark | DeepSeek-R1 | Grok-3 mini |
|---|---|---|
| Aider Polyglot | 71.4% | 49.3% |
| WeirdML | 41.6% | 42.6% |
| LMArena Coding | 1427 | 1380 |
| SciCode | 35.7% | — |
| LiveBench Coding | 66.7% | — |
| ALE-Bench | 804.12 | — |
| AlgoTune | 1.7 | — |
Agentic & Tool Use Not comparable
DeepSeek-R1: 30.4 (#68), Grok-3 mini: —
| Benchmark | DeepSeek-R1 | Grok-3 mini |
|---|---|---|
| DeepResearch Bench | 35.1% | — |
| BALROG | 34.9% | — |
| METR Time Horizons | 53.8% | — |
Reasoning DeepSeek-R1 leads
DeepSeek-R1: 18.7 (#253), Grok-3 mini: 13.8 (#323)
| Benchmark | DeepSeek-R1 | Grok-3 mini |
|---|---|---|
| ARC-AGI-2 | 1.3% | 0.4% |
| Kagi LLM Benchmark | 69.4% | 61.3% |
| ARC-AGI-1 | 21.2% | 16.5% |
| LMArena Hard Prompts | 1415 | 1374 |
| Epoch Capabilities Index | 141.3 | 140.35 |
| SimpleBench | 40.8% | — |
| CritPt | 1.1% | — |
| LiveBench Reasoning | 83.2% | — |
| LiveBench Data Analysis | 69.8% | — |
| ForecastBench | 60 | — |
| LiveBench | 71.6% | — |
Math DeepSeek-R1 leads
DeepSeek-R1: 42.0 (#70), Grok-3 mini: 40.3 (#78)
| Benchmark | DeepSeek-R1 | Grok-3 mini |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 66.4% | 77.8% |
| Omni-MATH | 42.4% | 31.8% |
| LMArena Math | 1401 | 1386 |
| MATH Level 5 | 96.6% | 90.9% |
| LiveBench Math | 80.7% | — |
| FrontierMath (Feb 2025 set) | — | 5.9% |
Knowledge Grok-3 mini leads
DeepSeek-R1: 44.3 (#81), Grok-3 mini: 46.3 (#75)
| Benchmark | DeepSeek-R1 | Grok-3 mini |
|---|---|---|
| GPQA Diamond | 76.3% | 76.3% |
| MMLU-Pro | 79.3% | 79.9% |
| Confabulations | 12.7% | 10.8% |
| GPQA (HELM) | 66.6% | 67.5% |
| LMArena Expert | 1394 | 1395 |
| Vectara Hallucination Rate | 11.3% | — |
Multilingual DeepSeek-R1 leads
DeepSeek-R1: 52.6 (#80), Grok-3 mini: 48.3 (#141)
| Benchmark | DeepSeek-R1 | Grok-3 mini |
|---|---|---|
| LMArena Non-English | 1412 | 1352 |
| LMArena Chinese | 1443 | 1387 |
| LMArena French | 1415 | 1357 |
| LMArena German | 1405 | 1350 |
| LMArena Japanese | 1394 | 1343 |
| LMArena Korean | 1359 | 1336 |
| LMArena Russian | 1423 | 1352 |
| LMArena Spanish | 1410 | 1379 |
Instruction Following Grok-3 mini leads
DeepSeek-R1: 72.2 (#138), Grok-3 mini: 78.7 (#8)
| Benchmark | DeepSeek-R1 | Grok-3 mini |
|---|---|---|
| IFEval | 78.4% | 95.1% |
| LMArena Instruction Following | 1382 | 1356 |
| LiveBench Instruction Following | 80.5% | — |
Long Context DeepSeek-R1 leads
DeepSeek-R1: 45.4 (#36), Grok-3 mini: 41.0 (#142)
| Benchmark | DeepSeek-R1 | Grok-3 mini |
|---|---|---|
| Fiction.LiveBench | 75% | 66.7% |
| LMArena Longer Query | 1391 | 1372 |
Writing & Preference DeepSeek-R1 leads
DeepSeek-R1: 61.5 (#84), Grok-3 mini: 52.7 (#163)
| Benchmark | DeepSeek-R1 | Grok-3 mini |
|---|---|---|
| LMArena Text | 1428 | 1370 |
| LMArena Creative Writing | 1406 | 1342 |
| Short-Story Creative Writing | 83% | 73.5% |
| WildBench | 82.8% | 65.1% |
| LMArena Multi-Turn | 1404 | 1355 |
| EQ-Bench Creative Writing | 1500 | — |
| LiveBench Language | 48.5% | — |
Frequently asked questions
Is DeepSeek-R1 better than Grok-3 mini?
DeepSeek-R1 is the stronger model overall, scoring 42.0 to 41.0 on the Noometry Index.
Is DeepSeek-R1 or Grok-3 mini better for coding?
DeepSeek-R1 scores higher on coding benchmarks: 45.9 versus 40.4 in the Noometry coding category.
How many benchmarks do DeepSeek-R1 and Grok-3 mini share?
34 benchmarks have published results for both models. DeepSeek-R1 has 52 scored results on Noometry and Grok-3 mini has 35.