Model comparison
DeepSeek V4.1 Flash vs Grok-3 mini
DeepSeek V4.1 Flash is the stronger model overall, scoring 52.8 to 41.2 on the Noometry Index.
Last verified . 20 shared benchmarks.
DeepSeekSummary
- They share 20 benchmarks with published results for both. DeepSeek V4.1 Flash scores higher in 7 categories and Grok-3 mini in 1 category; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where DeepSeek V4.1 Flash leads 50.2 to 13.6.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 98.3% for DeepSeek V4.1 Flash and 77.8% for Grok-3 mini.
- DeepSeek V4.1 Flash has downloadable open weights; the other is API-only.
Side by side
| DeepSeek V4.1 Flash | Grok-3 mini | |
|---|---|---|
| Provider | DeepSeek | xAI |
| Noometry Index | 52.8 | 41.2 |
| Released | 2026-09-09 | 2025-04-09 |
| Weights | Open | Proprietary |
| Context window | 1M | — |
| Max output | 393K | — |
| Input $ / M tokens | $0.15 | — |
| Output $ / M tokens | $0.60 | — |
| Results tracked | 37 | 35 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek V4.1 Flash leads
DeepSeek V4.1 Flash: 52.9 (#32), Grok-3 mini: 40.8 (#131)
| Benchmark | DeepSeek V4.1 Flash | Grok-3 mini |
|---|---|---|
| LMArena Coding | 1506 | 1379 |
| Aider Polyglot | — | 49.3% |
| LMArena WebDev | 1619 | — |
| SciCode | 51.9% | — |
| WeirdML | — | 42.6% |
| ALE-Bench | 1,092 | — |
Agentic & Tool Use Not comparable
DeepSeek V4.1 Flash: 31.2 (#69), Grok-3 mini: —
| Benchmark | DeepSeek V4.1 Flash | Grok-3 mini |
|---|---|---|
| APEX-Agents | 39.5% | — |
| GDP.pdf | 19.8% | — |
Reasoning DeepSeek V4.1 Flash leads
DeepSeek V4.1 Flash: 50.2 (#36), Grok-3 mini: 13.6 (#334)
| Benchmark | DeepSeek V4.1 Flash | Grok-3 mini |
|---|---|---|
| LMArena Hard Prompts | 1483 | 1375 |
| Epoch Capabilities Index | 154.9 | 140.35 |
| ARC-AGI-2 | — | 0.4% |
| Kagi LLM Benchmark | — | 61.3% |
| NYT Connections (extended) | 89.6% | — |
| ARC-AGI-1 | — | 16.5% |
| CritPt | 14.3% | — |
| Mystery Game Puzzles | 43% | — |
| DTBench | 89.9% | — |
| LMCA | 47% | — |
| Surface Evolver Bench | 46.3% | — |
Math DeepSeek V4.1 Flash leads
DeepSeek V4.1 Flash: 66.7 (#25), Grok-3 mini: 42.1 (#85)
| Benchmark | DeepSeek V4.1 Flash | Grok-3 mini |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 98.3% | 77.8% |
| LMArena Math | 1477 | 1386 |
| FrontierMath (Tiers 1-3) | 67.4% | — |
| FrontierMath Tier 4 | 26.8% | — |
| ProofBench | 54% | — |
| Omni-MATH | — | 31.8% |
| MATH Level 5 | — | 90.9% |
| FrontierMath (Feb 2025 set) | — | 5.9% |
Knowledge DeepSeek V4.1 Flash leads
DeepSeek V4.1 Flash: 57.9 (#38), Grok-3 mini: 46.4 (#81)
| Benchmark | DeepSeek V4.1 Flash | Grok-3 mini |
|---|---|---|
| GPQA Diamond | 89.8% | 76.3% |
| LMArena Expert | 1506 | 1395 |
| MMLU-Pro | — | 79.9% |
| Confabulations | — | 10.8% |
| GPQA (HELM) | — | 67.5% |
Multimodal Not comparable
DeepSeek V4.1 Flash: 39.1 (#61), Grok-3 mini: —
| Benchmark | DeepSeek V4.1 Flash | Grok-3 mini |
|---|---|---|
| LMArena Vision | 1277 | — |
| Furniture Assembly | 34.2% | — |
Multilingual DeepSeek V4.1 Flash leads
DeepSeek V4.1 Flash: 55.0 (#35), Grok-3 mini: 48.1 (#145)
| Benchmark | DeepSeek V4.1 Flash | Grok-3 mini |
|---|---|---|
| LMArena Non-English | 1448 | 1352 |
| LMArena Chinese | 1497 | 1387 |
| LMArena French | 1452 | 1357 |
| LMArena German | 1484 | 1349 |
| LMArena Japanese | 1412 | 1342 |
| LMArena Korean | 1452 | 1335 |
| LMArena Russian | 1471 | 1353 |
| LMArena Spanish | 1459 | 1381 |
Instruction Following Grok-3 mini leads
DeepSeek V4.1 Flash: 77.3 (#26), Grok-3 mini: 78.5 (#9)
| Benchmark | DeepSeek V4.1 Flash | Grok-3 mini |
|---|---|---|
| LMArena Instruction Following | 1474 | 1357 |
| IFEval | — | 95.1% |
Long Context DeepSeek V4.1 Flash leads
DeepSeek V4.1 Flash: 45.2 (#47), Grok-3 mini: 41.0 (#147)
| Benchmark | DeepSeek V4.1 Flash | Grok-3 mini |
|---|---|---|
| LMArena Longer Query | 1475 | 1372 |
| Fiction.LiveBench | — | 66.7% |
Writing & Preference DeepSeek V4.1 Flash leads
DeepSeek V4.1 Flash: 65.4 (#48), Grok-3 mini: 52.5 (#169)
| Benchmark | DeepSeek V4.1 Flash | Grok-3 mini |
|---|---|---|
| LMArena Text | 1462 | 1370 |
| LMArena Creative Writing | 1435 | 1342 |
| LMArena Multi-Turn | 1457 | 1355 |
| Short-Story Creative Writing | — | 73.5% |
| EQ-Bench Creative Writing | 1540 | — |
| WildBench | — | 65.1% |
Frequently asked questions
Is DeepSeek V4.1 Flash better than Grok-3 mini?
DeepSeek V4.1 Flash is the stronger model overall, scoring 52.8 to 41.2 on the Noometry Index.
Is DeepSeek V4.1 Flash or Grok-3 mini better for coding?
DeepSeek V4.1 Flash scores higher on coding benchmarks: 52.9 versus 40.8 in the Noometry coding category.
How many benchmarks do DeepSeek V4.1 Flash and Grok-3 mini share?
20 benchmarks have published results for both models. DeepSeek V4.1 Flash has 37 scored results on Noometry and Grok-3 mini has 35.