Model comparison
DeepSeek-V3.2-Exp vs Grok 4 Fast
DeepSeek-V3.2-Exp is the stronger model overall, scoring 44.3 to 39.4 on the Noometry Index.
Last verified . 26 shared benchmarks.
DeepSeekSummary
- They share 26 benchmarks with published results for both. DeepSeek-V3.2-Exp scores higher in 7 categories and Grok 4 Fast in 2 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where DeepSeek-V3.2-Exp leads 51.7 to 32.0.
- The biggest single-benchmark swing is Vectara Hallucination Rate: 5.3% for DeepSeek-V3.2-Exp and 19.7% for Grok 4 Fast.
- DeepSeek-V3.2-Exp has downloadable open weights; the other is API-only.
Side by side
| DeepSeek-V3.2-Exp | Grok 4 Fast | |
|---|---|---|
| Provider | DeepSeek | xAI |
| Noometry Index | 44.3 | 39.4 |
| Released | 2025-09-29 | 2025-09-19 |
| Weights | Open | Proprietary |
| Context window | 164K | — |
| Max output | 66K | — |
| Input $ / M tokens | $0.26 | — |
| Output $ / M tokens | $0.38 | — |
| Results tracked | 49 | 30 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek-V3.2-Exp leads
DeepSeek-V3.2-Exp: 46.5 (#65), Grok 4 Fast: 32.5 (#271)
| Benchmark | DeepSeek-V3.2-Exp | Grok 4 Fast |
|---|---|---|
| LMArena WebDev | 1362 | 1159 |
| WeirdML | 39.5% | 42.9% |
| LMArena Coding | 1454 | 1429 |
| SWE-bench Verified (bash only) | 70% | — |
| Aider Polyglot | 74.2% | — |
| SWE-bench Multilingual | 59% | — |
| SciCode | 38.9% | — |
Agentic & Tool Use DeepSeek-V3.2-Exp leads
DeepSeek-V3.2-Exp: 32.7 (#59), Grok 4 Fast: 29.5 (#86)
| Benchmark | DeepSeek-V3.2-Exp | Grok 4 Fast |
|---|---|---|
| Terminal-Bench | 39.6% | — |
| APEX-Agents | 21.3% | — |
| Berkeley Function Calling Leaderboard | 56.7% | — |
| TheAgentCompany | 42.9% | — |
| τ²-bench Banking | — | 15.7% |
| Cybench | — | 30% |
| LMArena Search | — | 1171 |
| Vending-Bench 2 | 1,034 | — |
Reasoning Too close to call
DeepSeek-V3.2-Exp: 22.1 (#208), Grok 4 Fast: 22.2 (#201)
| Benchmark | DeepSeek-V3.2-Exp | Grok 4 Fast |
|---|---|---|
| ARC-AGI-2 | 4% | 5.3% |
| Kagi LLM Benchmark | 52.2% | 66.1% |
| ARC-AGI-1 | 57% | 48.5% |
| LMArena Hard Prompts | 1434 | 1412 |
| DTBench | 87.7% | 82.7% |
| Epoch Capabilities Index | 146.27 | 144.2 |
| NYT Connections (extended) | 36.7% | — |
| CritPt | 2.9% | — |
| Chess Puzzles | 14% | — |
| Thematic Generalization | 65% | — |
| LMCA | 29.1% | — |
| ForecastBench | — | 60.5 |
Math DeepSeek-V3.2-Exp leads
DeepSeek-V3.2-Exp: 41.7 (#87), Grok 4 Fast: 38.9 (#123)
| Benchmark | DeepSeek-V3.2-Exp | Grok 4 Fast |
|---|---|---|
| LMArena Math | 1435 | 1419 |
| MathArena Final-Answer Competitions | 57.7% | — |
| OTIS Mock AIME 2024-2025 | 87.8% | — |
| ProofBench | 8% | — |
| FrontierMath (Feb 2025 set) | 22.1% | — |
| FrontierMath Tier 4 (v1) | 2.1% | — |
Knowledge DeepSeek-V3.2-Exp leads
DeepSeek-V3.2-Exp: 51.7 (#66), Grok 4 Fast: 32.0 (#214)
| Benchmark | DeepSeek-V3.2-Exp | Grok 4 Fast |
|---|---|---|
| Vectara Hallucination Rate | 5.3% | 19.7% |
| LMArena Expert | 1436 | 1411 |
| GPQA Diamond | 83.4% | — |
Multilingual Too close to call
DeepSeek-V3.2-Exp: 52.2 (#90), Grok 4 Fast: 51.3 (#111)
| Benchmark | DeepSeek-V3.2-Exp | Grok 4 Fast |
|---|---|---|
| LMArena Non-English | 1409 | 1396 |
| LMArena Chinese | 1461 | 1457 |
| LMArena French | 1433 | 1431 |
| LMArena German | 1440 | 1383 |
| LMArena Japanese | 1374 | 1352 |
| LMArena Korean | 1371 | 1357 |
| LMArena Russian | 1424 | 1389 |
| LMArena Spanish | 1440 | 1415 |
Instruction Following DeepSeek-V3.2-Exp leads
DeepSeek-V3.2-Exp: 74.5 (#93), Grok 4 Fast: 73.2 (#121)
| Benchmark | DeepSeek-V3.2-Exp | Grok 4 Fast |
|---|---|---|
| LMArena Instruction Following | 1413 | 1387 |
Long Context Grok 4 Fast leads
DeepSeek-V3.2-Exp: 47.6 (#16), Grok 4 Fast: 63.2 (#3)
| Benchmark | DeepSeek-V3.2-Exp | Grok 4 Fast |
|---|---|---|
| Fiction.LiveBench | 83.3% | 94.4% |
| LMArena Longer Query | 1428 | 1415 |
| CL-bench | 13.2% | — |
| CL-bench Life | 9.5% | — |
Writing & Preference DeepSeek-V3.2-Exp leads
DeepSeek-V3.2-Exp: 62.4 (#77), Grok 4 Fast: 60.0 (#102)
| Benchmark | DeepSeek-V3.2-Exp | Grok 4 Fast |
|---|---|---|
| LMArena Text | 1425 | 1407 |
| LMArena Creative Writing | 1403 | 1387 |
| LMArena Multi-Turn | 1427 | 1414 |
| EQ-Bench Creative Writing | 1515 | — |
Frequently asked questions
Is DeepSeek-V3.2-Exp better than Grok 4 Fast?
DeepSeek-V3.2-Exp is the stronger model overall, scoring 44.3 to 39.4 on the Noometry Index.
Is DeepSeek-V3.2-Exp or Grok 4 Fast better for coding?
DeepSeek-V3.2-Exp scores higher on coding benchmarks: 46.5 versus 32.5 in the Noometry coding category.
How many benchmarks do DeepSeek-V3.2-Exp and Grok 4 Fast share?
26 benchmarks have published results for both models. DeepSeek-V3.2-Exp has 49 scored results on Noometry and Grok 4 Fast has 30.