Model comparison
DeepSeek-V3.2-Exp vs Grok-3 mini
DeepSeek-V3.2-Exp is the stronger model overall, scoring 44.3 to 41.2 on the Noometry Index.
Last verified . 27 shared benchmarks.
DeepSeekSummary
- They share 27 benchmarks with published results for both. DeepSeek-V3.2-Exp scores higher in 6 categories and Grok-3 mini in 2 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where DeepSeek-V3.2-Exp leads 62.4 to 52.5.
- The biggest single-benchmark swing is ARC-AGI-1: 57% for DeepSeek-V3.2-Exp and 16.5% for Grok-3 mini.
- DeepSeek-V3.2-Exp has downloadable open weights; the other is API-only.
Side by side
| DeepSeek-V3.2-Exp | Grok-3 mini | |
|---|---|---|
| Provider | DeepSeek | xAI |
| Noometry Index | 44.3 | 41.2 |
| Released | 2025-09-29 | 2025-04-09 |
| Weights | Open | Proprietary |
| Context window | 164K | — |
| Max output | 66K | — |
| Input $ / M tokens | $0.26 | — |
| Output $ / M tokens | $0.38 | — |
| Results tracked | 49 | 35 |
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-3 mini: 40.8 (#131)
| Benchmark | DeepSeek-V3.2-Exp | Grok-3 mini |
|---|---|---|
| Aider Polyglot | 74.2% | 49.3% |
| WeirdML | 39.5% | 42.6% |
| LMArena Coding | 1454 | 1379 |
| SWE-bench Verified (bash only) | 70% | — |
| LMArena WebDev | 1362 | — |
| SWE-bench Multilingual | 59% | — |
| SciCode | 38.9% | — |
Agentic & Tool Use Not comparable
DeepSeek-V3.2-Exp: 32.7 (#59), Grok-3 mini: —
| Benchmark | DeepSeek-V3.2-Exp | Grok-3 mini |
|---|---|---|
| Terminal-Bench | 39.6% | — |
| APEX-Agents | 21.3% | — |
| Berkeley Function Calling Leaderboard | 56.7% | — |
| TheAgentCompany | 42.9% | — |
| Vending-Bench 2 | 1,034 | — |
Reasoning DeepSeek-V3.2-Exp leads
DeepSeek-V3.2-Exp: 22.1 (#208), Grok-3 mini: 13.6 (#334)
| Benchmark | DeepSeek-V3.2-Exp | Grok-3 mini |
|---|---|---|
| ARC-AGI-2 | 4% | 0.4% |
| Kagi LLM Benchmark | 52.2% | 61.3% |
| ARC-AGI-1 | 57% | 16.5% |
| LMArena Hard Prompts | 1434 | 1375 |
| Epoch Capabilities Index | 146.27 | 140.35 |
| NYT Connections (extended) | 36.7% | — |
| CritPt | 2.9% | — |
| Chess Puzzles | 14% | — |
| Thematic Generalization | 65% | — |
| DTBench | 87.7% | — |
| LMCA | 29.1% | — |
Math Too close to call
DeepSeek-V3.2-Exp: 41.7 (#87), Grok-3 mini: 42.1 (#85)
| Benchmark | DeepSeek-V3.2-Exp | Grok-3 mini |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 87.8% | 77.8% |
| LMArena Math | 1435 | 1386 |
| FrontierMath (Feb 2025 set) | 22.1% | 5.9% |
| MathArena Final-Answer Competitions | 57.7% | — |
| ProofBench | 8% | — |
| Omni-MATH | — | 31.8% |
| MATH Level 5 | — | 90.9% |
| FrontierMath Tier 4 (v1) | 2.1% | — |
Knowledge DeepSeek-V3.2-Exp leads
DeepSeek-V3.2-Exp: 51.7 (#66), Grok-3 mini: 46.4 (#81)
| Benchmark | DeepSeek-V3.2-Exp | Grok-3 mini |
|---|---|---|
| GPQA Diamond | 83.4% | 76.3% |
| LMArena Expert | 1436 | 1395 |
| MMLU-Pro | — | 79.9% |
| Confabulations | — | 10.8% |
| Vectara Hallucination Rate | 5.3% | — |
| GPQA (HELM) | — | 67.5% |
Multilingual DeepSeek-V3.2-Exp leads
DeepSeek-V3.2-Exp: 52.2 (#90), Grok-3 mini: 48.1 (#145)
| Benchmark | DeepSeek-V3.2-Exp | Grok-3 mini |
|---|---|---|
| LMArena Non-English | 1409 | 1352 |
| LMArena Chinese | 1461 | 1387 |
| LMArena French | 1433 | 1357 |
| LMArena German | 1440 | 1349 |
| LMArena Japanese | 1374 | 1342 |
| LMArena Korean | 1371 | 1335 |
| LMArena Russian | 1424 | 1353 |
| LMArena Spanish | 1440 | 1381 |
Instruction Following Grok-3 mini leads
DeepSeek-V3.2-Exp: 74.5 (#93), Grok-3 mini: 78.5 (#9)
| Benchmark | DeepSeek-V3.2-Exp | Grok-3 mini |
|---|---|---|
| LMArena Instruction Following | 1413 | 1357 |
| IFEval | — | 95.1% |
Long Context DeepSeek-V3.2-Exp leads
DeepSeek-V3.2-Exp: 47.6 (#16), Grok-3 mini: 41.0 (#147)
| Benchmark | DeepSeek-V3.2-Exp | Grok-3 mini |
|---|---|---|
| Fiction.LiveBench | 83.3% | 66.7% |
| LMArena Longer Query | 1428 | 1372 |
| CL-bench | 13.2% | — |
| CL-bench Life | 9.5% | — |
Writing & Preference DeepSeek-V3.2-Exp leads
DeepSeek-V3.2-Exp: 62.4 (#77), Grok-3 mini: 52.5 (#169)
| Benchmark | DeepSeek-V3.2-Exp | Grok-3 mini |
|---|---|---|
| LMArena Text | 1425 | 1370 |
| LMArena Creative Writing | 1403 | 1342 |
| LMArena Multi-Turn | 1427 | 1355 |
| Short-Story Creative Writing | — | 73.5% |
| EQ-Bench Creative Writing | 1515 | — |
| WildBench | — | 65.1% |
Frequently asked questions
Is DeepSeek-V3.2-Exp better than Grok-3 mini?
DeepSeek-V3.2-Exp is the stronger model overall, scoring 44.3 to 41.2 on the Noometry Index.
Is DeepSeek-V3.2-Exp or Grok-3 mini better for coding?
DeepSeek-V3.2-Exp scores higher on coding benchmarks: 46.5 versus 40.8 in the Noometry coding category.
How many benchmarks do DeepSeek-V3.2-Exp and Grok-3 mini share?
27 benchmarks have published results for both models. DeepSeek-V3.2-Exp has 49 scored results on Noometry and Grok-3 mini has 35.