Model comparison
DeepSeek V4 Pro vs Grok 4.1 Fast
DeepSeek V4 Pro is the stronger model overall, scoring 54.3 to 41.4 on the Noometry Index. Grok 4.1 Fast costs 3.6× less per token, which makes it the better buy when DeepSeek V4 Pro's lead doesn't matter for your workload.
Last verified . 27 shared benchmarks.
DeepSeekSummary
- They share 27 benchmarks with published results for both. DeepSeek V4 Pro scores higher in 8 categories and Grok 4.1 Fast in 1 category; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where DeepSeek V4 Pro leads 64.8 to 31.9.
- The biggest single-benchmark swing is ProofBench: 50% for DeepSeek V4 Pro and 4% for Grok 4.1 Fast.
- Grok 4.1 Fast is cheaper at $0.20 / $0.50 per million input/output tokens, against $0.66 / $1.98 for DeepSeek V4 Pro.
- DeepSeek V4 Pro accepts more context: 1M tokens versus 128K.
- DeepSeek V4 Pro has downloadable open weights; the other is API-only.
Side by side
| DeepSeek V4 Pro | Grok 4.1 Fast | |
|---|---|---|
| Provider | DeepSeek | xAI |
| Noometry Index | 54.3 | 41.4 |
| Released | 2026-04-24 | 2025-06-27 |
| Weights | Open | Proprietary |
| Context window | 1M | 128K |
| Max output | 393K | 30K |
| Input $ / M tokens | $0.66 | $0.20 |
| Output $ / M tokens | $1.98 | $0.50 |
| Results tracked | 48 | 32 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek V4 Pro leads
DeepSeek V4 Pro: 52.4 (#34), Grok 4.1 Fast: 34.1 (#245)
| Benchmark | DeepSeek V4 Pro | Grok 4.1 Fast |
|---|---|---|
| LMArena WebDev | 1582 | 1242 |
| LMArena Coding | 1470 | 1411 |
| ALE-Bench | 1,403 | 394.93 |
| SWE-bench Verified | 77.6% | — |
| FrontierCode | 28.6% | — |
| SciCode | 51% | — |
| WeirdML | 66.2% | — |
Agentic & Tool Use Grok 4.1 Fast leads
DeepSeek V4 Pro: 32.8 (#58), Grok 4.1 Fast: 36.3 (#39)
| Benchmark | DeepSeek V4 Pro | Grok 4.1 Fast |
|---|---|---|
| Vending-Bench 2 | 3,285 | 1,107 |
| APEX-Agents | 47.3% | — |
| Berkeley Function Calling Leaderboard | — | 69.6% |
| τ²-bench Banking | — | 13.1% |
| LMArena Search | — | 1171 |
Reasoning DeepSeek V4 Pro leads
DeepSeek V4 Pro: 56.5 (#24), Grok 4.1 Fast: 43.4 (#49)
| Benchmark | DeepSeek V4 Pro | Grok 4.1 Fast |
|---|---|---|
| NYT Connections (extended) | 91.3% | 87.4% |
| LMArena Hard Prompts | 1461 | 1407 |
| DTBench | 93.9% | 87.7% |
| ForecastBench | 56.1 | 61 |
| ARC-AGI-2 | 61.3% | — |
| SimpleBench | — | 56% |
| Kagi LLM Benchmark | 53.5% | — |
| ARC-AGI-1 | 90.5% | — |
| CritPt | 18% | — |
| Chess Puzzles | 47% | — |
| Mystery Game Puzzles | 43% | — |
| LMCA | 45.5% | — |
| Surface Evolver Bench | 40% | — |
| Epoch Capabilities Index | 155.31 | — |
Math DeepSeek V4 Pro leads
DeepSeek V4 Pro: 64.8 (#30), Grok 4.1 Fast: 31.9 (#221)
| Benchmark | DeepSeek V4 Pro | Grok 4.1 Fast |
|---|---|---|
| MathArena Final-Answer Competitions | 76.6% | 60.9% |
| ProofBench | 50% | 4% |
| LMArena Math | 1455 | 1408 |
| FrontierMath (Tiers 1-3) | 64.6% | — |
| FrontierMath Tier 4 | 26.8% | — |
| OTIS Mock AIME 2024-2025 | 98.6% | — |
Knowledge DeepSeek V4 Pro leads
DeepSeek V4 Pro: 59.5 (#31), Grok 4.1 Fast: 33.1 (#207)
| Benchmark | DeepSeek V4 Pro | Grok 4.1 Fast |
|---|---|---|
| Vectara Hallucination Rate | 8.6% | 17.8% |
| LMArena Expert | 1464 | 1399 |
| GPQA Diamond | 91.7% | — |
| SimpleQA Verified | 52.9% | — |
Multimodal Not comparable
DeepSeek V4 Pro: —, Grok 4.1 Fast: 37.0 (#76)
| Benchmark | DeepSeek V4 Pro | Grok 4.1 Fast |
|---|---|---|
| LMArena Vision | — | 1201 |
Multilingual DeepSeek V4 Pro leads
DeepSeek V4 Pro: 54.4 (#45), Grok 4.1 Fast: 51.0 (#114)
| Benchmark | DeepSeek V4 Pro | Grok 4.1 Fast |
|---|---|---|
| LMArena Non-English | 1439 | 1391 |
| LMArena Chinese | 1486 | 1441 |
| LMArena French | 1472 | 1415 |
| LMArena German | 1458 | 1404 |
| LMArena Japanese | 1445 | 1349 |
| LMArena Korean | 1447 | 1361 |
| LMArena Russian | 1453 | 1387 |
| LMArena Spanish | 1458 | 1413 |
Instruction Following DeepSeek V4 Pro leads
DeepSeek V4 Pro: 76.1 (#47), Grok 4.1 Fast: 72.7 (#133)
| Benchmark | DeepSeek V4 Pro | Grok 4.1 Fast |
|---|---|---|
| LMArena Instruction Following | 1448 | 1376 |
Long Context DeepSeek V4 Pro leads
DeepSeek V4 Pro: 45.0 (#51), Grok 4.1 Fast: 42.4 (#126)
| Benchmark | DeepSeek V4 Pro | Grok 4.1 Fast |
|---|---|---|
| LMArena Longer Query | 1458 | 1390 |
| CL-bench Life | 13.5% | — |
Writing & Preference DeepSeek V4 Pro leads
DeepSeek V4 Pro: 65.5 (#46), Grok 4.1 Fast: 57.2 (#131)
| Benchmark | DeepSeek V4 Pro | Grok 4.1 Fast |
|---|---|---|
| LMArena Text | 1451 | 1408 |
| LMArena Creative Writing | 1446 | 1394 |
| EQ-Bench Creative Writing | 1553 | 1327 |
| LMArena Multi-Turn | 1467 | 1389 |
| EQ-Bench 4 | 1166 | — |
Frequently asked questions
Is DeepSeek V4 Pro better than Grok 4.1 Fast?
DeepSeek V4 Pro is the stronger model overall, scoring 54.3 to 41.4 on the Noometry Index. Grok 4.1 Fast costs 3.6× less per token, which makes it the better buy when DeepSeek V4 Pro's lead doesn't matter for your workload.
Which is cheaper, DeepSeek V4 Pro or Grok 4.1 Fast?
Grok 4.1 Fast is cheaper. It lists at $0.20 per million input tokens and $0.50 per million output tokens; DeepSeek V4 Pro lists at $0.66 and $1.98.
Is DeepSeek V4 Pro or Grok 4.1 Fast better for coding?
DeepSeek V4 Pro scores higher on coding benchmarks: 52.4 versus 34.1 in the Noometry coding category.
Which has the bigger context window?
DeepSeek V4 Pro does, with 1M tokens against 128K.
How many benchmarks do DeepSeek V4 Pro and Grok 4.1 Fast share?
27 benchmarks have published results for both models. DeepSeek V4 Pro has 48 scored results on Noometry and Grok 4.1 Fast has 32.