Model comparison
DeepSeek V4 Flash vs Grok 4
DeepSeek V4 Flash is the stronger model overall, scoring 53.6 to 48.1 on the Noometry Index.
Last verified . 26 shared benchmarks.
DeepSeekSummary
- They share 26 benchmarks with published results for both. DeepSeek V4 Flash scores higher in 5 categories and Grok 4 in 3 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in long context, where Grok 4 leads 63.1 to 43.8.
- The biggest single-benchmark swing is ARC-AGI-2: 61.4% for DeepSeek V4 Flash and 16% for Grok 4.
- DeepSeek V4 Flash has downloadable open weights; the other is API-only.
Side by side
| DeepSeek V4 Flash | Grok 4 | |
|---|---|---|
| Provider | DeepSeek | xAI |
| Noometry Index | 53.6 | 48.1 |
| Released | 2026-04-24 | 2025-07-09 |
| Weights | Open | Proprietary |
| Context window | 1M | — |
| Max output | 393K | — |
| Input $ / M tokens | $0.15 | — |
| Output $ / M tokens | $0.60 | — |
| Results tracked | 41 | 48 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Grok 4 leads
DeepSeek V4 Flash: 47.9 (#59), Grok 4: 50.3 (#46)
| Benchmark | DeepSeek V4 Flash | Grok 4 |
|---|---|---|
| WeirdML | 63% | 45.7% |
| LMArena Coding | 1457 | 1408 |
| FrontierCode | 18.8% | — |
| Aider Polyglot | — | 79.6% |
| LMArena WebDev | 1582 | — |
| SciCode | 49.9% | — |
| ALE-Bench | 1,306 | — |
Agentic & Tool Use Not comparable
DeepSeek V4 Flash: —, Grok 4: 32.3 (#68)
| Benchmark | DeepSeek V4 Flash | Grok 4 |
|---|---|---|
| Terminal-Bench | — | 27.2% |
| Berkeley Function Calling Leaderboard | — | 63% |
| GDPval | — | 21.1% |
| Cybench | — | 43% |
| DeepResearch Bench | — | 47.3% |
| BALROG | — | 43.6% |
| LMArena Search | — | 1142 |
| METR Time Horizons | — | 66.6% |
Reasoning DeepSeek V4 Flash leads
DeepSeek V4 Flash: 53.7 (#30), Grok 4: 36.7 (#65)
| Benchmark | DeepSeek V4 Flash | Grok 4 |
|---|---|---|
| ARC-AGI-2 | 61.4% | 16% |
| SimpleBench | 61.1% | 60.5% |
| Kagi LLM Benchmark | 52.2% | 73.6% |
| ARC-AGI-1 | 89% | 66.7% |
| Chess Puzzles | 33% | 28% |
| LMArena Hard Prompts | 1444 | 1409 |
| Epoch Capabilities Index | 154.49 | 146.44 |
| NYT Connections (extended) | 89.6% | — |
| CritPt | 16.6% | — |
| Mystery Game Puzzles | 34% | — |
| DTBench | 90.9% | — |
| LMCA | 41.7% | — |
| ForecastBench | — | 60.9 |
Math DeepSeek V4 Flash leads
DeepSeek V4 Flash: 60.3 (#37), Grok 4: 48.4 (#64)
| Benchmark | DeepSeek V4 Flash | Grok 4 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 94.4% | 84% |
| LMArena Math | 1427 | 1422 |
| FrontierMath (Tiers 1-3) | 57.5% | — |
| FrontierMath Tier 4 | 24.4% | — |
| MathArena Final-Answer Competitions | 76.5% | — |
| ProofBench | 56% | — |
| Omni-MATH | — | 60.3% |
| FrontierMath (Feb 2025 set) | — | 19.7% |
| FrontierMath Tier 4 (v1) | — | 2.1% |
Knowledge DeepSeek V4 Flash leads
DeepSeek V4 Flash: 55.4 (#48), Grok 4: 53.8 (#55)
| Benchmark | DeepSeek V4 Flash | Grok 4 |
|---|---|---|
| GPQA Diamond | 91% | 87% |
| LMArena Expert | 1441 | 1415 |
| SimpleQA Verified | 33.6% | — |
| MMLU-Pro | — | 85.1% |
| Confabulations | — | 12.4% |
| GPQA (HELM) | — | 72.7% |
Multimodal Not comparable
DeepSeek V4 Flash: —, Grok 4: 33.7 (#94)
| Benchmark | DeepSeek V4 Flash | Grok 4 |
|---|---|---|
| LMArena Vision | — | 1210 |
| GeoBench | — | 45% |
Multilingual DeepSeek V4 Flash leads
DeepSeek V4 Flash: 53.0 (#72), Grok 4: 51.8 (#103)
| Benchmark | DeepSeek V4 Flash | Grok 4 |
|---|---|---|
| LMArena Non-English | 1420 | 1403 |
| LMArena Chinese | 1468 | 1427 |
| LMArena French | 1439 | 1418 |
| LMArena German | 1418 | 1429 |
| LMArena Japanese | 1406 | 1394 |
| LMArena Korean | 1384 | 1377 |
| LMArena Russian | 1428 | 1410 |
| LMArena Spanish | 1436 | 1420 |
Instruction Following Grok 4 leads
DeepSeek V4 Flash: 74.9 (#81), Grok 4: 79.2 (#5)
| Benchmark | DeepSeek V4 Flash | Grok 4 |
|---|---|---|
| LMArena Instruction Following | 1421 | 1387 |
| IFEval | — | 94.9% |
Long Context Grok 4 leads
DeepSeek V4 Flash: 43.8 (#85), Grok 4: 63.1 (#4)
| Benchmark | DeepSeek V4 Flash | Grok 4 |
|---|---|---|
| LMArena Longer Query | 1434 | 1409 |
| Fiction.LiveBench | — | 94.4% |
Writing & Preference DeepSeek V4 Flash leads
DeepSeek V4 Flash: 63.8 (#61), Grok 4: 58.5 (#116)
| Benchmark | DeepSeek V4 Flash | Grok 4 |
|---|---|---|
| LMArena Text | 1432 | 1411 |
| LMArena Creative Writing | 1403 | 1397 |
| LMArena Multi-Turn | 1449 | 1416 |
| Short-Story Creative Writing | — | 76.9% |
| EQ-Bench Creative Writing | 1559 | — |
| WildBench | — | 79.7% |
Frequently asked questions
Is DeepSeek V4 Flash better than Grok 4?
DeepSeek V4 Flash is the stronger model overall, scoring 53.6 to 48.1 on the Noometry Index.
Is DeepSeek V4 Flash or Grok 4 better for coding?
Grok 4 scores higher on coding benchmarks: 50.3 versus 47.9 in the Noometry coding category.
How many benchmarks do DeepSeek V4 Flash and Grok 4 share?
26 benchmarks have published results for both models. DeepSeek V4 Flash has 41 scored results on Noometry and Grok 4 has 48.