Model comparison
DeepSeek-V3 vs Grok 3
DeepSeek-V3 and Grok 3 score almost the same on the Noometry Index (39.2 vs 39.6), so choose on price, context window or the category you care about most.
Last verified . 36 shared benchmarks.
DeepSeekSummary
- They share 36 benchmarks with published results for both. DeepSeek-V3 scores higher in 3 categories and Grok 3 in 5 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Grok 3 leads 46.0 to 37.5.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 37.8% for DeepSeek-V3 and 55.6% for Grok 3.
- DeepSeek-V3 has downloadable open weights; the other is API-only.
Side by side
| DeepSeek-V3 | Grok 3 | |
|---|---|---|
| Provider | DeepSeek | xAI |
| Noometry Index | 39.2 | 39.6 |
| Released | 2024-12-26 | 2025-04-09 |
| Weights | Open | Proprietary |
| Context window | 164K | — |
| Max output | 164K | — |
| Input $ / M tokens | $0.24 | — |
| Output $ / M tokens | $0.90 | — |
| Results tracked | 60 | 40 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
DeepSeek-V3: 41.9 (#99), Grok 3: 41.5 (#113)
| Benchmark | DeepSeek-V3 | Grok 3 |
|---|---|---|
| Aider Polyglot | 55.1% | 53.3% |
| WeirdML | 36.1% | 37.2% |
| LMArena Coding | 1368 | 1432 |
| SciCode | 35.8% | — |
| BigCodeBench Instruct | 50% | — |
| LiveBench Coding | 70.9% | — |
| BigCodeBench Complete | 62.2% | — |
| HumanEval+ | 86.6% | — |
| MBPP+ | 73% | — |
Agentic & Tool Use Not comparable
DeepSeek-V3: —, Grok 3: 30.2 (#69)
| Benchmark | DeepSeek-V3 | Grok 3 |
|---|---|---|
| BALROG | — | 29.5% |
| METR Time Horizons | 49.6% | — |
Reasoning DeepSeek-V3 leads
DeepSeek-V3: 20.0 (#226), Grok 3: 13.8 (#322)
| Benchmark | DeepSeek-V3 | Grok 3 |
|---|---|---|
| SimpleBench | 27.2% | 36.1% |
| Kagi LLM Benchmark | 52.3% | 61.3% |
| LMArena Hard Prompts | 1365 | 1433 |
| Epoch Capabilities Index | 135.95 | 138.34 |
| ARC-AGI-2 | — | 0% |
| ARC-AGI-1 | — | 5.5% |
| CritPt | 0% | — |
| LiveBench Reasoning | 65.8% | — |
| DTBench | 64.8% | — |
| LiveBench Data Analysis | 60.9% | — |
| LMCA | 15.5% | — |
| BIG-Bench Hard | 87.5% | — |
| ForecastBench | 59.1 | — |
| HellaSwag | 88.9% | — |
| LiveBench | 66.9% | — |
| PIQA | 84.7% | — |
| WinoGrande | 85.2% | — |
Math Grok 3 leads
DeepSeek-V3: 30.4 (#208), Grok 3: 36.2 (#122)
| Benchmark | DeepSeek-V3 | Grok 3 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 37.8% | 55.6% |
| Omni-MATH | 40.3% | 46.4% |
| LMArena Math | 1373 | 1390 |
| MATH Level 5 | 75.5% | 88.7% |
| FrontierMath (Feb 2025 set) | 1.7% | 3.8% |
| LiveBench Math | 73.5% | — |
| FrontierMath Tier 4 (v1) | — | 0% |
Knowledge Grok 3 leads
DeepSeek-V3: 37.5 (#147), Grok 3: 46.0 (#76)
| Benchmark | DeepSeek-V3 | Grok 3 |
|---|---|---|
| GPQA Diamond | 67.6% | 75.8% |
| MMLU-Pro | 72.3% | 78.8% |
| Confabulations | 26.1% | 14.2% |
| Vectara Hallucination Rate | 6.1% | 5.8% |
| GPQA (HELM) | 53.8% | 65% |
| LMArena Expert | 1352 | 1420 |
| ARC (AI2) Challenge | 95.3% | — |
| MMLU | 87.2% | — |
| TriviaQA | 82.9% | — |
Multilingual Grok 3 leads
DeepSeek-V3: 48.8 (#138), Grok 3: 52.5 (#81)
| Benchmark | DeepSeek-V3 | Grok 3 |
|---|---|---|
| LMArena Non-English | 1358 | 1411 |
| LMArena Chinese | 1390 | 1448 |
| LMArena French | 1386 | 1460 |
| LMArena German | 1374 | 1431 |
| LMArena Japanese | 1333 | 1389 |
| LMArena Korean | 1321 | 1374 |
| LMArena Russian | 1372 | 1416 |
| LMArena Spanish | 1358 | 1417 |
Instruction Following Grok 3 leads
DeepSeek-V3: 73.0 (#125), Grok 3: 75.2 (#69)
| Benchmark | DeepSeek-V3 | Grok 3 |
|---|---|---|
| IFEval | 83.2% | 88.4% |
| LMArena Instruction Following | 1345 | 1409 |
| LiveBench Instruction Following | 81.5% | — |
Long Context Grok 3 leads
DeepSeek-V3: 34.0 (#248), Grok 3: 38.7 (#187)
| Benchmark | DeepSeek-V3 | Grok 3 |
|---|---|---|
| Fiction.LiveBench | 50% | 58.3% |
| LMArena Longer Query | 1353 | 1439 |
Writing & Preference DeepSeek-V3 leads
DeepSeek-V3: 57.6 (#124), Grok 3: 56.0 (#136)
| Benchmark | DeepSeek-V3 | Grok 3 |
|---|---|---|
| LMArena Text | 1375 | 1426 |
| LMArena Creative Writing | 1364 | 1414 |
| Short-Story Creative Writing | 77% | 76.4% |
| EQ-Bench Creative Writing | 1472 | 1186 |
| WildBench | 83% | 84.9% |
| LMArena Multi-Turn | 1389 | 1424 |
| LiveBench Language | 49.1% | — |
Frequently asked questions
Is DeepSeek-V3 better than Grok 3?
DeepSeek-V3 and Grok 3 score almost the same on the Noometry Index (39.2 vs 39.6), so choose on price, context window or the category you care about most.
Is DeepSeek-V3 or Grok 3 better for coding?
They score almost the same on coding (41.9 vs 41.5); test both on your own repository before choosing.
How many benchmarks do DeepSeek-V3 and Grok 3 share?
36 benchmarks have published results for both models. DeepSeek-V3 has 60 scored results on Noometry and Grok 3 has 40.